Brown&twomey etal 2011 revision of ranitomeya by Emy J. Riquero

ISSN 1175-5326 (print edition)

Monograph

ZOOTAXA ISSN 1175-5334 (online edition)

ZOOTAXA 3083

A taxonomic revision of the Neotropical poison frog genus Ranitomeya (Amphibia: Dendrobatidae) JASON L. BROWN1*, EVAN TWOMEY2, ADOLFO AMÉZQUITA3, MOISÉS BARBOSA DE SOUZA4, JANALEE P. CALDWELL5, STEFAN LÖTTERS6, RUDOLF VON MAY7, PAULO ROBERTO MELO-SAMPAIO4, DANIEL MEJÍA-VARGAS8, PEDRO PEREZ-PEÑA9, MARK PEPPER10, ERIK H. POELMAN11, MANUEL SANCHEZ-RODRIGUEZ12 & KYLE SUMMERS2 1. Biology Department, Duke University, Durham NC 27701, USA 2. Biology Department, East Carolina University, Greenville NC 27858, USA 3. Universidad de los Andes, Department of Biological Sciences, Bogotá, Colombia 4. Universidade Federal do Acre, Centro de Ciências Biológicas e da Natureza, BR 364 km 04 Distrito Industrial, CEP 69.915-900 Rio Branco AC, Brazil 5. Sam Noble Oklahoma Museum of Natural History and Department of Zoology, University of Oklahoma, Norman, OK 73072, USA 6. Biogeography Department, Trier University, 54286 Trier, Germany 7. Department of Biological Sciences, Florida International University, Miami FL 33199, USA 8. Associate Researcher, Department of Biological Sciences, Universidad de los Andes, Bogotá, Colombia 9. Wildlife Conservation Society, Malecón Tarapaca 332, Iquitos Perú. 10. Understory Enterprises Inc., Charing Cross, Ontario Canada, N0P 1G0 11. Laboratory of Entomology, Wageningen University, P.O. Box 8031, 6700 EH Wageningen, the Netherlands 12. Understory Enterprises Inc., Iquitos, Peru *jasonleebrown@gmail.com

Magnolia Press Auckland, New Zealand

Accepted by M. Vences: 15 Sep. 2011; published: 28 Oct. 2011

Jason L. Brown, Evan Twomey, Adolfo Amézquita, Moisés Barbosa de Souza, Janalee P. Caldwell, Stefan Lötters, Rudolf von May, Paulo Roberto Melo-Sampaio, Daniel Mejía-Vargas, Pedro PerezPeña, Mark Pepper, Erik H. Poelman, Manuel Sanchez-Rodriguez & Kyle Summers A taxonomic revision of the Neotropical poison frog genus Ranitomeya (Amphibia: Dendrobatidae) (Zootaxa 3083) 120 pp.; 30 cm. 28 Oct. 2011 ISBN 978-1-86977-815-6 (paperback) ISBN 978-1-86977-816-3 (Online edition)

FIRST PUBLISHED IN 2011 BY Magnolia Press P.O. Box 41-383 Auckland 1346 New Zealand e-mail: zootaxa@mapress.com http://www.mapress.com/zootaxa/

© 2011 Magnolia Press All rights reserved. No part of this publication may be reproduced, stored, transmitted or disseminated, in any form, or by any means, without prior written permission from the publisher, to whom all requests to reproduce copyright material should be directed in writing. This authorization does not extend to any other kind of copying, by any means, in any form, and for any purpose

Please other than private research use. note that we intended to have the running ISSN 1175-5326 ISSN 1175-5334

title "BROWN & TWOMEY ET AL", given Evan's tremendous contributions to this manuscript. (Print edition) However for some reason post-proof it was (Online edition) changed to “BROWN ET AL”. Thanks- Evan for all your contributions. The manuscript certainly benefited greatly for your contributions. -Jason L. Brown

BROWN ET AL.

Table of contents Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Resumen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Resumo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Résumé . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Difficulties of Ranitomeya taxonomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 History of Ranitomeya . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Goals of this paper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Materials and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Key to the ‘small’ aposematic poison frogs of South America . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Systematics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Andinobates Twomey, Brown, Amézquita & Mejía-Vargas, gen. nov. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Species groups contained within Andinobates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Andinobates minutus species group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Andinobates fulguritus species group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Andinobates bombetes group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 Ranitomeya Bauer 1988 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Species groups contained within Ranitomeya . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 Ranitomeya defleri species group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 Ranitomeya toraro Brown, Caldwell, Twomey, Melo-Sampaio & Souza sp. nov. . . . . . . . . . 42 Ranitomeya defleri Twomey & Brown 2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Ranitomeya reticulata species group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 Ranitomeya benedicta Brown, Twomey, Pepper and Sanchez-Rodriguez 2008 . . . . . . . . . . . 48 Ranitomeya fantastica Boulenger 1884 “1883” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 Ranitomeya reticulata Boulenger 1884 “1883” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 Ranitomeya summersi Brown, Twomey, Pepper & Sanchez-Rodriguez 2008 . . . . . . . . . . . . 54 Ranitomeya uakarii Brown, Schulte & Summers 2006 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 Ranitomeya ventrimaculata Shreve 1935 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Ranitomeya vanzolinii species group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 Ranitomeya cyanovittata Perez-Peña, Chavez, Brown & Twomey 2010 . . . . . . . . . . . . . . . . 61 Ranitomeya flavovittata Schulte 1999 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 Ranitomeya imitator Schulte 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 Ranitomeya sirensis Aichinger 1991 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 Ranitomeya vanzolinii Myers 1982 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 Ranitomeya yavaricola Perez-Peña, Chavez, Brown & Twomey 2010 . . . . . . . . . . . . . . . . . 78 Ranitomeya variabilis species group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 Ranitomeya amazonica Schulte 1999 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 Ranitomeya variabilis Zimmermann & Zimmermann 1988 . . . . . . . . . . . . . . . . . . . . . . . . . . 86 Incertae sedis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 Ranitomeya rubrocephala Schulte 1999 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 Phylogenetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 Notes on taxonomic changes and the use of the name Ranitomeya . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 Tadpole morphology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 Müllerian mimicry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 Dendrobatid frogs and the pet trade . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 Atalaya, Ucayali, Peru . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 Tingo Maria, Huánuco, Peru . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 Future directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 DNA sequences: Incertae sedis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 Appendix I. Species names, collection localities, and GenBank accession numbers (where available) for taxa included . . . . . . . . . 118 Appendix II. Material examined and current taxonomic status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119

TAXONOMIC REVISION OF RANITOMEYA

Abstract The Neotropical poison frog genus Ranitomeya is revised, resulting in one new genus, one new species, five synonymies and one species classified as nomen dubium. We present an expanded molecular phylogeny that contains 235 terminals, 104 of which are new to this study. Notable additions to this phylogeny include seven of the 12 species in the minuta group, 15 Ranitomeya amazonica, 20 R. lamasi, two R. sirensis, 30 R. ventrimaculata and seven R. uakarii. Previous researchers have long recognized two distinct, reciprocally monophyletic species groups contained within Ranitomeya, sensu Grant et al. 2006: the ventrimaculata group, which is distributed throughout much of the Amazon, and the minuta group of the northern Andes and Central America. We restrict Ranitomeya to the former group and erect a new genus, Andinobates Twomey, Brown, Amézquita & Mejía-Vargas gen. nov., for members of the minuta group. Other major taxonomic results of the current revision include the following: (i) A new species, Ranitomeya toraro Brown, Caldwell, Twomey, Melo-Sampaio & Souza sp. nov., is described from western Brazil. This species has long been referred to as R. ventrimaculata but new morphological and phylogenetic data place it sister to R. defleri. (ii) Examination of the holotype of R. ventrimaculata revealed that this specimen is in fact a member of what is currently referred to as R. duellmani, therefore, Dendrobates duellmani Schulte 1999 is considered herein a junior synonym of D. ventrimaculatus Shreve 1935 (= R. ventrimaculata). (iii) For the frogs that were being called R. ventrimaculata prior to this revision, the oldest available and therefore applicable name is R. variabilis. Whereas previous definitions of R. variabilis were restricted to spotted highland frogs near Tarapoto, Peru, our data suggest that this color morph is conspecific with lowland striped counterparts. Therefore, the definition of R. variabilis is greatly expanded to include most frogs which were (prior to this revision) referred to as R. ventrimaculata. (iv) Phylogenetic and bioacoustic evidence support the retention of R. amazonica as a valid species related to R. variabilis as defined in this paper. Based on phylogenetic data, R. amazonica appears to be distributed throughout much of the lower Amazon, as far east as French Guiana and the Amazon Delta and as far west as Iquitos, Peru. (v) Behavioral and morphological data, as well as phylogenetic data which includes topotypic material of R. sirensis and numerous samples of R. lamasi, suggest that the names sirensis, lamasi and biolat are applicable to a single, widespread species that displays considerable morphological variation throughout its range. The oldest available name for this group is sirensis Aichinger; therefore, we expand the definition of R. sirensis. (vi) Ranitomeya ignea and R. intermedia, elevated to the species status in a previous revision, are placed as junior synonyms of R. reticulata and R. imitator, respectively. (vii) Ranitomeya rubrocephala is designated as nomen dubium. In addition to taxonomic changes, this revision includes the following: (i) Explicit definitions of species groups that are consistent with our proposed taxonomy. (ii) A comprehensive dichotomous key for identification of ‘small’ aposematic poison frogs of South and Central America. (iii) Detailed distribution maps of all Ranitomeya species, including unpublished localities for most species. In some cases, these records result in substantial range extensions (e.g., R. uakarii, R. fantastica). (iv) Tadpole descriptions for R. amazonica, R. flavovittata, R. imitator, R. toraro sp. nov., R. uakarii and R. variabilis; plus a summary of tadpole morphological data for Andinobates and Ranitomeya species. (v) A summary of call data on most members of Andinobates and Ranitomeya, including call data of several species that have not been published before. (vi) A discussion on the continued impacts of the pet trade on poison frogs (vii) A discussion on several cases of potential Müllerian mimicry within the genus Ranitomeya. We also give opinions regarding the current debate on recent taxonomic changes and the use of the name Ranitomeya. Key words: Andinobates gen. nov., Dendrobates, Ranitomeya toraro sp. nov., systematics, Müllerian mimicry, phylogenetics

Resumen Se hace una revisión del género de ranas venenosas neotropicales Ranitomeya, con el resultado de un nuevo género, una nueva especie, cinco sinonimias y una especie clasificada como nomen dubium. Presentamos una filogenia molecular amplia que contiene 235 terminales, de los cuales 104 son nuevos para este estudio. Las adiciones más importantes en este estudio incluyen a siete de las 12 especies del grupo minuta, 15 de Ranitomeya amazonica, 20 de R. lamasi, dos de R. sirensis, 30 de R. ventrimaculata y siete de R. uakarii. Los pasados estudios han distinguido por un largo tiempo dos grupos de especies distintos, recíprocamente monofíléticos dentro del género Ranitomeya, sensu Grant et al. 2006: El grupo ventrimaculata, distribuido a lo largo de gran parte de la amazonia y el grupo minuta, del norte de los Andes y Centroamérica. Restringimos Ranitomeya al grupo ventrimaculata y erigimos un nuevo género, Andinobates Twomey, Brown, Amézquita & Mejía-Vargas gen. nov., para los miembros del grupo minuta. Se incluyen además los siguientes resultados taxonómicos mayores en esta revisión: (i) Se describe una nueva especie, Ranitomeya toraro Brown, Caldwell, Twomey, Melo-Sampaio & Souza sp. nov., del occidente de Brasil. Esta especie ha sido por largo tiempo conocida como R. ventrimaculata, pero nuevos datos morfológicos y filogenéticos la sitúan como la especie hermana de R. defleri. (ii). El examen del holotípo de R. ventrimaculata reveló que este especímen es, de hecho, un miembro de lo que actualmente se conoce como R. duellmani, por lo que Dendrobates du-

BROWN ET AL.

ellmani Schulte 1999 se considera aquí como un sinónimo junior de D. ventrimaculatus Shreve 1935 (= R. ventrimaculata). (iii) Para las ranas que estaban siendo llamadas R. ventrimaculata antes de esta revisión, el nombre más antiguo y por tanto aplicable es R. variabilis. Mientras que las definiciones anteriores de R. variabilis la restringían a las ranas con puntos de tierras altas cerca a Tarapoto, nuestros datos sugieren que estas ranas son conespecíficas a sus contrapartes con líneas de tierras bajas. Por lo tanto, la definición de R. variabilis se expande ampliamente y se sugiere incluir muchas de las ranas que eran (anteriormente a esta revisión) referidas a R. ventrimaculata. (iv) La evidencia filogenética y bioacustica soporta la conservación de R. amazonica como una especie válida, aunque relacionada a otras. Basado en evidencia filogenética, R. amazonica parece distribuirse a lo largo de mucho del bajo Amazonas, de Guyana Francesa y el delta del Amazonas al occidente y desde Iquitos, Perú al occidente. (v) La evidencia comportamental y morfológica, aunada a nuestros resultados de análisis filogenéticos, que incluyen material topotípico de R. sirensis, así como numerosas muestras de R. lamasi, sugieren que los nombres sirensis, lamasi y biolat son aplicables a una única especie, ampliamente extendida, que muestra una variación morfológica considerable a lo largo de su distribución. El nombre más antiguo disponible para este grupo es sirensis Aichinger; por lo tanto, ampliamos la definición de R. sirensis. (vi) Ranitomeya ignea y R. intermedia, que fueron elevadas al nivel de especie en una revisión previa, son puestas como sinónimos junior de R. reticulata y R. imitator respectivamente. (vii) Se designa a Ranitomeya rubrocephala como nomen dubium. En adición a los cambios taxonómicos, esta revisión incluye lo siguiente: (i) Definiciones explícitas de los grupos de especies que son consistentes con nuestra taxonomía propuesta. (ii) Una clave dicotómica completa para la identificación de ranas aposemáticas "pequeñas" de sur y centro América. (iii) Mapas de distribución detallados para todas las especies de Ranitomeya, que incluyen localidades no publicadas para muchas especies. En algunos casos, estos reportes proporcionan extensiones de rango considerables (por ejemplo R. uakarii y R. fantastica). (iv) Descripciones de los renacuajos de R. amazonica, R. flavovittata, R. imitator, R. toraro Brown, Caldwell, Twomey, Melo-Sampaio & Souza sp. nov., R. uakarii y R. variabilis; además de un compendio de los datos de morfología larvaria para las especies de Andinobates y Ranitomeya. (v) Una recopilación de los cantos de muchos de los miembros de Andinobates y Ranitomeya, que incluye los cantos de muchas especies que no habían sido publicadas. (vi) Una discusión sobre los impactos continuos del comercio de animales en las ranas venenosas. (vii) Introducimos varios casos potenciales de mímica Mülleriana dentro del género Ranitomeya. Además, opinamos acerca del debate actual sobre los recientes cambios taxonómicos y el uso del nombre Ranitomeya.

Resumo Os sapos venenosos neotropicais do gênero Ranitomeya são revisados aqui, resultando em um novo gênero, uma nova espécie, cinco sinonímias e uma das espécies classificadas como nomen dubium. Apresentamos uma filogenia molecular expandida que contém 235 terminais, 104 dos quais são novos para este estudo. Incluem a essa filogenia adições notáveis para sete das 12 espécies do grupo minuta, 15 Ranitomeya amazonica, 20 R. lamasi, dois R. sirensis, 30 R. ventrimaculata e sete R. uakarii. Outros pesquisadores já haviam reconhecido por muito tempo dois grupos reciprocamente distintos, de espécies monofiléticas contidas em Ranitomeya sensu Grant et al. 2006: o grupo ventrimaculata, que é distribuído pela maior parte da Amazônia, e o grupo minuta do norte dos Andes e da América Central. Nós restringimos Ranitomeya ao grupo ventrimaculata e elevamos um novo gênero, Andinobates Twomey, Brown, Amézquita & Mejía-Vargas gen. nov., para os membros do grupo minuta. Outros importantes resultados taxonômicos da revisão em curso incluem: (i) Uma nova espécie, Ranitomeya toraro Brown, Caldwell, Twomey, Melo-Sampaio & Souza sp. nov., é descrita para o Brasil ocidental. Esta espécie tem sido referida como R. ventrimaculata, mas novos dados morfológicos e filogenéticos colocam-na como espécie-irmã de R. defleri. (ii) O exame do holótipo de R. ventrimaculata revelou que o espécime é na verdade um membro do que é atualmente designado como Dendrobates.duellmani Schulte 1999, portanto, este é considerado aqui como sinônimo júnior de D. ventrimaculatus Shreve 1935. (iii) Para os anuros que estavam sendo chamados R. ventrimaculata antes desta revisão, o nome mais antigo disponível e, portanto, aplicável é R.variabilis. Considerando que as definições anteriores de R. variabilis eram restritas aos sapos manchados de altitude próxima a Tarapoto, Peru, nossos dados sugerem que estes sapos são da mesma espécie que seus homólogos listrados da planície. Portanto, a definição de R. variabilis é expandida e sugerimos incluir a maioria dos anuros que estavam (antes desta revisão) designados como R. ventrimaculata. (iv) Evidências bioacústicas e filogenéticas apoiam a manutenção de R. amazonica como uma espécie válida, mas com espécies relacionadas. Baseado em dados filogenéticos, R. amazonica parece ser distribuída em grande parte do baixo Amazonas, a leste, até a Guiana Francesa e do delta do Amazonas e no extremo oeste, até Iquitos, no Peru. (v) Dados morfológicos, comportamentais e análises filogenéticas dos nossos resultados, que incluem material topotípico de R. sirensis bem como numerosas amostras de R. lamasi, sugerem que os nomes sirensis, lamasi e biolat são aplicáveis a uma única espécie que apresenta ampla variação morfológica em toda a sua dsitribuição. O nome mais antigo disponível para este grupo é Ranitomeya sirensis Aichinger, portanto, podemos expandir a definição de R.sirensis. (vi) Ranitomeya ignea e R. intermedia, elevadas ao status de espécies em uma revisão anterior, são colocadas como sinônimos júnior de R. reticulata e R.imitator, respectivamente. (vii) Ranitomeya rubrocephala é designada como nomen dubium. TAXONOMIC REVISION OF RANITOMEYA

Além de mudanças taxonômicas, esta revisão inclui: (i) definições de grupos de espécies que são consistentes com a nossa taxonomia proposta. (ii) Uma chave para a identificação completa de 'pequenos' sapos venenosos aposemáticos da América Central e do Sul. (iii) Mapas de distribuição detalhada de todas as espécies de Ranitomeya, incluindo locais inéditos para a maioria das espécies. Em alguns casos, esses registros resultam no conjunto de extensões substanciais (para R. uakarii, R. fantastica por exemplo). (iv) Descrição dos girinos de R. amazonica, R. flavovittata, R. imitator, R. toraro sp. nov., R. uakarii e R. variabilis acrescido de um resumo dos dados morfológicos do girino para as espécie de Andinobates e Ranitomeya. (v) Um resumo dos dados das vocalizações da maioria dos membros Andinobates e Ranitomeya incluindo dados de vocalização de várias espécies que não tinham sido publicadas antes. (vi) Uma discussão sobre os impactos da continuidade do comércio de sapos venenosos como animais de estimação (vii) Introduzimos vários casos de potencial mimetismo Mülleriano dentro do gênero Ranitomeya. Também opinamos a respeito do debate atual sobre as recentes mudanças taxonômicas e o uso do nome Ranitomeya.

Résumé La révision de la systématique des grenouilles vénéneuses du genre Ranitomeya résulte en un nouveau genre, une nouvelle espèce, cinq synonymes, et une espèce classifiée en tant que nomen dubium. Nous présentons une phylogénie moléculaire élargie contenant 235 terminaux, dont 104 sont exclusifs à cette étude. Parmi les ajouts notables à cette phylogénie il y a l’inclusion de sept des 12 espèces du groupe minuta, 15 Ranitomeya amazonica, 20 R. lamasi, deux R. sirensis, 30 R. ventrimaculata et sept R. uakarii. Les scientifiques reconnaissent depuis longtemps deux groupes d’espèces distincts, caractérisés par leur monophylie réciproque, au sein du genre Ranitomeya sensu Grant et al. 2006: le groupe ventrimaculata distribué en Amazonie, et le groupe minuta du nord des Andes et de l’Amérique Centrale. Nous restreignons le genre Ranitomeya au groupe ventrimaculata et créons un nouveau genre, Twomey, Brown, Amézquita & Mejía-Vargas gen. nov., pour les membres du groupe minuta. Parmi les changements taxonomiques majeurs découlant de la présente révision, il est à souligner : (i) Une nouvelle espèce, Ranitomeya toraro Brown, Caldwell, Twomey, Melo-Sampaio & Souza sp. nov., provenant de l’ouest du Brésil est décrite. Cette espèce a longtemps été définie comme étant R. ventrimaculata mais de nouvelles données morphologiques et phylogénétiques la situent comme étant une espèce sœur de R. defleri. (ii) La réexamination de l’holotype de R. ventrimaculata a révélé que ce spécimen fait parti du groupe connu comme étant R. duellmani, et de fait, Dendrobates duellmani Schulte 1999 est considéré comme étant un synonyme de D. ventrimaculatus Shreve 1935. (iii) Pour les grenouilles appelées R. ventrimaculata avant cette révision, le plus vieux nom disponible, et de ce fait applicable, est R. variabilis. Alors que R. variabilis était antérieurement caractérisé comme étant une grenouille réticulée provenant des hautes altitudes près de Tarapoto, nos données suggèrent que ces grenouilles sont conspécifiques avec leurs congénères lignés et distribués dans les basses terres. Ainsi, la définition de R. variabilis est grandement élargie, suggérant que la majorité des grenouilles définies (avant cette étude) en tant que R. ventrimculata sont en fait des R. variabilis. (iv) Les données phylogénétiques et bioacoustiques maintiennent R. amazonica comme étant une espèce distincte, proche de R. variabilis tel que défini dans cet article. Selon ces données phylogénétiques, la distribution de R. amazonica semble s’étendre sur la majorité de l’Amazonie : de la Guyane Française et du delta du fleuve Amazone, jusqu’à Iquitos au Pérou. (v) Des données comportementales, morphologiques, ainsi que nos analyses phylogénétiques, incluant du matériel topotypique de R. sirensis ainsi que de nombreux échantillons de R. lamasi, suggèrent que les noms sirensis, lamasi et biolat sont applicables à une seule et même espèce ayant une grande distribution et démontrant de grandes variations morphologiques. Le nom le plus ancien pour ce groupe est sirensis Aichinger; nous élargissons donc la définition de R. sirensis. (vi) Ranitomeya ignea et R. intermedia, élevés au statut d’espèce dans une révision antérieure, sont classés comme étant des synonymes juniors de R. reticulata et R. imitator respectivement. (vii) Ranitomeya rubrocephala est désigné nomen dubium. En plus des changements taxonomiques, cette révision inclut : (i) Des définitions explicites des différents groupes d’espèces définis dans la taxonomie proposée. (ii) Une clé dichotomique permettant l’identification de ces ‘petites’ grenouilles aposématiques d’Amérique Central et du Sud. (iii) Des cartes de distributions détaillées pour les différentes espèces de Ranitomeya, incluant de nouvelles localités pour la majorité des espèces. Dans certains cas, l’aire de répartition est substantiellement élargie (e.g., R. uakarii, R. fantastica). (iv) La description des têtards de R. amazonica, R. flavovittata, R. imitator, R. toraro sp. nov., R. uakarii et R. variabilis; ainsi qu’un résumé des données morphologiques pour les espèces d’Andinobates et Ranitomeya. (v) Un résumé des données de chants pour la majorité des membres des genres Andinobates et Ranitomeya, incluant des données jamais publiées auparavant pour plusieurs espèces. (vi) Une discussion de l’impact qu’a le braconnage des grenouilles vénéneuses qui sont destinées au marché animalier. (viii) La mention de plusieurs cas potentiels de mimétisme Müllerien au sein du genre Ranitomeya. Nous donnons également notre opinion en ce qui concerne le débat sur les changements taxonomiques récents et de l’utilisation du nom Ranitomeya.

BROWN ET AL.

Introduction Difficulties of Ranitomeya taxonomy In the early 1970s, as a graduate student, Philip A. Silverstone sorted through countless jars of small Neotropical frogs with the daunting task of clarifying the taxonomy of aposematic poison dart frogs (family Dendrobatidae). Many of the frogs he worked with and allocated to the genus Dendrobates are now considered to be members of the more recently erected genus Ranitomeya, a species-rich group of poison frogs characterized by their diminutive size. The initial classifications of members of Ranitomeya, without genetic, behavioral, or acoustic data, would have been extremely difficult. Therefore, it is easy to understand why early workers were careful and ‘lumped’ every small Amazonian specimen with pale limb reticulation under the single name Dendrobates quinquevittatus, which was suggested to be a highly variable species that exhibited a cline of morphology (see Fig. 14 in Silverstone 1975). Such levels of variation were certainly not unheard of in poison frogs. For example, Oophaga pumilio, O. histrionica, Dendrobates tinctorius, and D. auratus display dramatic levels of phenotypic variation over their geographic ranges— the alpha-taxonomic status of each of these species continues to be widely accepted today (e.g., Lötters et al. 2007). In 1982, Myers wrote that: New material, as well as closer attention to morphological details and evidence of sympatry, convinces me that Dendrobates quinquevittatus Steindachner, sensu Silverstone, is a composite of five or more species of distinctively colored frogs. C.W. Myers 1982, p. 3 Interestingly, almost a decade later, Caldwell & Myers (1990) would point out that, in fact, none of those species was actually Dendrobates quinquevittatus sensu stricto. They stated that: Myers erred in concluding that the restricted quinquevittatus remained “indeed a widespread and variable species.” We correct that mistake by further restricting the name Dendrobates quinquevittatus Steindachner to a not particularly variable species confined to the southwestern section of Amazonia and by formally resurrecting the name Dendrobates ventrimaculatus Shreve for the residue of what recently has been called quinquevittatus. We emphasize that Dendrobates ventrimaculatus sensu lato becomes the new composite species of this complex, but we urge caution in dealing with it systematically. We are aware that “ventrimaculatus” contains more than one species, some of which may be difficult to separate even when sympatric…Attempts at taxonomic diagnosis based on literature and/or terrarium specimens of vague provenance seem likely to be fruitless at best… or, more seriously, to create needless confusion arising from nomenclatural irresponsibility. J.P. Caldwell & C.W. Myers 1990, p. 11 Today, 13 species are recognized that were considered by Silverstone (1975) as D. quinquevittatus. Dendrobates quinquevittatus is just one example of how this group has perplexed taxonomists for the last half-century, and, despite the fact that many researchers have suspected that this group was composed of many more species (e.g., Myers 1982; Caldwell & Myers 1990; Noonan & Wray 2006; Brown et al. 2006; Twomey & Brown 2009), defining species boundaries has proven to be extremely difficult. This confusion, naturally, slowed the rate at which species have been described (Fig. 1). Furthermore, this group’s variation appeared to be limitless, with every unexplored valley or mountain seeming to contain new forms, many times muddling the distinction between previously distinct forms. Christmann (2004) lamented “From excursion to excursion, our desire to see the largest number of populations in their biotype increases. However, at the same time—with each new find, our feelings of helplessness grow. Will it ever be possible to recognize some kind of order in this multitude of species?” Myers (1982) commented “But we are gradually realizing that there always will be another poison frog to be sought on the next mountain or across the next river. So, the new names are offered in hope of encouraging others who find it interesting to pursue small frogs in obscure places.” The vital piece of the quinquevittatus puzzle that was previously lacking was not morphological data (of which there was plenty), but acoustic, genetic and behavioral data, and firsthand experience with these species in their natural habitats throughout their entire ranges. TAXONOMIC REVISION OF RANITOMEYA

FIGURE 1. Accumulation of described species in genera Ranitomeya and Andinobates.

This last point is particularly crucial because most of the alpha-taxonomic problems in Ranitomeya stem from the fact that many species could not be diagnosed with certainty based on coloration (or, at least, reliable characters were not apparent), and other diagnostic morphological characters (e.g., osteology) seemed to be insufficiently variable for diagnoses below the genus level. Many species show extensive variation between populations (e.g., R. imitator, R. fantastica, R. amazonica), whereas other species display only subtle differences when compared to their sister taxa (e.g., R. flavovittata and R. vanzolinii). One cause for the varying levels of intraspecific diversity in Amazonian species (and resulting taxonomic confusion), is due to at least one, and perhaps multiple, occurrences of Müllerian mimicry (see Symula et al. 2001 and the mimicry section in discussion of this paper). For example, in some localities in north-central Peru, three species have converged (at least partially) on a single phenotype (i.e., Ranitomeya imitator, R. variabilis sensu this paper and R. fantastica near Pongo de Cainarachi, Department San Martin), making it easier for local predators to recognize the similarly toxic species (Ruxton et al. 2004). On the other hand, nearby populations of the same species appear so dissimilar (having converged on the color and pattern of a different model species) that they are not easily identified as conspecific unless acoustic and genetic data are available. This group’s taxonomic classification was further complicated by the shy nature of many species and their typically high species diversity at any given site, with two to three species usually occurring in sympatry, often appearing similar in color and pattern. Furthermore, in many species, advertisement calls can be easily overlooked, often being quiet, infrequent, and insect-like. Nevertheless, advertisement calls are important characters in the classification of the main groups within Ranitomeya (see Table 5; Erdtmann & Amézquita 2009). In addition, the size of a species’ distribution seems to bear little significance in determining specific units, and some species appear to have extremely small ranges (e.g., R. summersi), whereas other species appeared to be distributed across much of the Amazon, from French Guiana to Peru (e.g., R. amazonica and R. uakarii). Recently, the use of molecular phylogenetics has elucidated many of these problems and in some instances exposed flaws in previously accepted taxonomies (see for example the species accounts of R. ventrimaculata, R. variabilis, R. sirensis in this paper). Using these methods, researchers have been able to focus their taxonomic studies on closely related species groups, rather than higher-level groups that may have been easier to diagnose based on morphology.

History of Ranitomeya George A. Boulenger, more than a century ago, described the first species in this group, Dendrobates fantasticus and D. reticulatus (Boulenger 1884 "1883"). In 1975, Philip A. Silverstone published the first comprehensive monograph dealing with specific relationships of the genus Dendrobates (a group containing Ranitomeya sensu Grant et al. 2006). Using adult and tadpole morphology he proposed the minutus group, containing six species:

BROWN ET AL.

altobueyensis, fulguritus, minutus, opisthomelas, quinquevittatus and steyermarki. The main characters used to define this group were larval: emarginate (laterally indented) oral disc and dextral anus (though he lacked data for quinquevittatus and steyermarki). Soon after, Charles W. Myers and John W. Daly began publishing their seminal research on the systematics of Central and South American dendrobatids. Their addition of alkaloid profiles, vocalization data and behavioral data to traditional taxonomic methods clarified many of the coarse relationships within this group and led to the description of numerous new species. In 1980, they described Dendrobates bombetes and resurrected D. reticulatus, assigning them and an undescribed species that was eventually named D. claudiae by Jungfer et al. (2000) to Silverstone’s minutus group. Furthermore, they hypothesized that abditus, bombetes and opisthomelas formed a monophyletic group delimited by ‘median gap that interrupts the papillate fringe on the posterior (lower) edge of the oral disc’ (Myers & Daly 1980). Shortly after, Myers removed D. fantasticus from synonymy with D. quinquevittatus and placed D. vanzolinii, D. fantasticus, D. captivus, D. quinquevittatus, and D. reticulatus (removing the latter two from the minutus group) in a suggested monophyletic assemblage delimited by ‘distinctively reticulated limbs,’ which was dubbed the quinquevittatus group (Myers 1982). Five years later, Myers placed the remaining members of the minutus group into a new genus, Minyobates (with steyermarki as the type species), based on the following characters: cephalic amplexus present, very small size (12–19.5 mm snout to vent length), oblique lateral stripe present (though he qualified this by noting that this feature is also absent in some Minyobates), and larvae with lateral indentations of the oral disc of the mouth and a dextral anus (Myers 1987). The establishment of a new genus was hurried to preceed the expected name Ranitomeya, which shortly after was coined by Luc Bauer, a Dutch amateur herpetologist, in a privately published paper (Bauer 1988). Ranitomeya at that time contained the former Dendrobates species reticulatus (as the type species), captivus, fantasticus, imitator, mysteriosus, quinquevittatus and vanzolinii, and was completely ignored in the scientific literature, presumably because it posed several problems under the International Code of Zoological Nomenclature (ICZN, see Grant et al. 2006 for a detailed discussion). Caldwell & Myers (1990) further restricted D. quinquevittatus and the quinquevittatus group to D. quinquevittatus sensu stricto and its suggested sister taxon D. castaneoticus, united by their shared absence of the inner metacarpal tubercle. They placed this group as sister to the clade united by pale limb reticulation (i.e., Ranitomeya of Grant et al. 2006). This clade was referred to as the ventrimaculatus group, and it comprised all other members of the quinquevittatus group exclusive of quinquevittatus sensu stricto, but the authors did not propose additional synapomorphies to support this arrangement. More than a decade later, Rainer Schulte privately published a book on Peruvian dendrobatids in which he described five new species and proposed several novel systematic relationships, including the designation of eight species groups (Schulte 1999). His classifications did not consider much of the works of previous authors (not allowing him to classify all taxa in previously established groups) and some proposed groups were paraphyletic in the illustrated diagrams. More importantly, synapomorphies and the criteria from which the groups were designated were largely not mentioned. A fair scientific assessment of these groups cannot be made and these groups will not be considered herein (see Table 1); for a detailed discussion of Schulte’s work see Lötters & Vences (2000). In the late 1990s, Kyle Summers and collaborators started using molecular phylogenetics to look at the specific relationships among a few members of the genus Ranitomeya (Summers et al. 1997). Since, numerous studies have addressed the molecular phylogenetics of this group (e.g., Summers et al. 1999; Clough & Summers 2000; Vences et al. 2000; Symula et al. 2001, 2003; Santos et al. 2003, 2009; Darst & Cannatella 2004; Brown et al. 2006, 2008c; Grant et al. 2006; Noonan & Wray 2006; Roberts et al. 2006a; Twomey & Brown 2008). Using primarily molecular phylogenetics, Grant et al. (2006) revised the entire family Dendrobatidae, dividing Dendrobates sensu Silverstone (1975) into five genera: Adelphobates, Dendrobates, Minyobates, Oophaga, and Ranitomeya. Grant et al. (2006) placed the former members of the minutus group sensu Myers (1982), minus Minyobates steyermarki, as well as most members of the ventrimaculatus group sensu Caldwell & Myers (1990), minus captivus and mysteriosus, in the genus Ranitomeya (an available name under the ICZN). Only steyermarki was retained in Minyobates. These placements were well supported by prior phylogenies (Vences et al. 2000, 2003; Noonan & Wray 2006; Roberts et al. 2006a). The definition of Ranitomeya contained 24 species (Table 1) and is currently the most widely accepted taxonomy for this group (though see below). In 2008, Twomey & Brown described a new genus, Excidobates, removing captivus from Adelphobates and placing it in its own group together with mysteriosus (an affinity originally suggested by Myers 1982), which was sister to the genus Ranitomeya (Twomey & Brown 2008). They also suggested that future taxonomies may be better off by restricting the use of

TAXONOMIC REVISION OF RANITOMEYA

Ranitomeya to the ventrimaculata group sensu Caldwell & Myers (1990) and placing the minuta group into a separate, at that time, unnamed genus. Roberts et al. (2006a) and Twomey & Brown (2008), collectively, further subdivided the ventrimaculata group into three smaller groups: reticulata, vanzolinii and ventrimaculata (Table 1), in part to help accommodate the numerous new species being described within this genus. More recently, in online supplemental material, Santos et al. (2009) proposed to synonymize Adelphobates, Dendrobates, Excidobates, Minyobates, Oophaga and Ranitomeya back into Dendrobates. We will use here the previously described taxonomy following Grant et al. (2006) and Twomey & Brown (2008). For further comments on this matter see discussion. The genus Ranitomeya currently comprises two reciprocally monophyletic clades: the minuta group and a clade containing the ventrimaculata, reticulata and vanzolinii groups. The minuta group is trans-Andean; its members have long, rattle-like advertisement calls and typically possess dark limbs and coarsely blotched venters. The ventrimaculata, reticulata and vanzolinii groups comprise a motley clade of Amazonian frogs, most of which possess pale limb reticulation and some permutation of dorsolateral stripes.

Goals of This Paper Based on results from our molecular phylogenetic analyses, accompanied by morphological (adult and larval), behavioral and acoustic data, and natural history information, we: (i) restrict Ranitomeya to the ventrimaculata, reticulata and vanzolinii groups and redefine (in one case rename) these groups and erect a fourth species group. We also reevaluate the taxonomic status of all taxa contained within this genus, resulting in the description of one new species and the redefinitions of others, as well as several synonymies. (ii) describe a new genus, Andinobates, for species in the minuta group. At the same time we restrict the minuta group to claudiae and minuta and erect two species groups for the remaining species in the new genus.

Materials and methods Species examination. Measurements were made with mechanical calipers and a dissecting microscope to the nearest 0.01 mm following Myers (1982) and Brown et al. (2006): snout to vent length (SVL), femur length from vent to lateral surface of knee (FL), tibia length from heel to lateral surface of knee (TL), knee–knee distance with both thighs extended at a 90 degree angle from body (respective to each side), forming a straight line between both (KK), foot length from proximal edge of metatarsal tubercle to tip of toe IV (FoL), hand length from proximal edge of metacarpal tubercle to tip of longest finger (HaL), head length from most exposed corner of occipitum to tip of snout (HL), head width between tympanum (HW), body width under axilla (BW), upper eyelid width (UEW), interorbital distance (IOD), horizontal tympanum diameter (TD), horizontal eye diameter (ED), distance from outer corner of eye to tympanum (DET), length of finger I from proximal edge of median palmar tubercle to tip of finger disc (L1F), length of finger II from proximal edge of median palmar tubercle to tip of finger disc (L2F), width of disc of finger III (W3D) and width of finger III just below disc (W3F). All tadpole measurements are in mm. All SVL measurements are based on adult individuals unless stated otherwise. All voucher specimens morphologically examined in this paper were deposited in the Museu Paraense Emílio Goeldi and Ministério da Ciência e Tecnologia do Brasil, Belém, Pará, Brazil (MPEG) and in the Sam Noble Oklahoma Museum of Natural History, Norman, OK, USA (OMNH). As follows are abbreviations of museums mentioned: California Academy of Sciences, San Francisco, CA, USA (CAS); Museum of Comparative Zoology, Harvard University, Cambridge, MA, USA (MCZ); Smithsonian National Museum of Natural History, Washington, DC, USA (USNM); Museo de Historia Natural ‘Javier Prado’ de la Universidad Nacional Mayor de San Marcos, Lima, Peru (MUSM); Natural History Museum, London, United Kingdom (NHML); Muséum d'histoire naturelle de la Ville de Genève, Geneva, Switzerland (MHNG); Natural History Museum, University of Kansas, Lawrence, KS, USA (KU); Centro de Ornitología y Biodiversidad, Surco-Lima, Peru (CORBIDI); Rainer Schulte Private Collection, Tarapoto, Peru (CRS); Naturhistorisches Museum Wien, Vienna, Austria (NHMW); Museum of Zoology of the University of São Paulo, São Paulo, Brazil (MZUSP); Museo de la Facultad de Ciencias Biológicas

BROWN ET AL.

de la Universidad Nacional de la Amazonía Peruana, Iquitos, Peru (MZUNAP); Museum national d'histoire naturelle, Paris, France (MNHN); Los Angeles County Museum of Natural History, Los Angeles, CA, USA (LACM); Naturhistoriska Riksmuseet, Göteborg, Sweden (NHMG); Staatliches Museum für Naturkunde Stuttgart, Stuttgart, Germany (SMNS); Pontificia Universidad Javeriana in Bogotá, Colombia (PUJB). Tadpoles (n = 5) of R. toraro sp. nov were collected from the backs of male parent frogs, preserved in 10% formalin for description and staged according to Gosner (1960). All other tadpoles were from the following sources: the backs of male parent frogs or offspring of captive frogs from known localities (typically from parents that are F2s; the tadpoles were either preserved, as described above, or measured alive). Tadpoles that were measured alive were anesthetized by placing a small drop of benzocaine on the top of head and allowed to recover after measurements. The Labial Tooth Row Formula (LTRF) follows Altig & McDiarmid (1999). Adult morphological characters used herein are the same as used in Grant et al. (2006) with the exception that we classify the stripe that extends longitudinally down the back (above the vertebral column) as the middorsal stripe versus medial stripe in Grant et al. (2006), see Figure 2. An estimate of conservation priority based on IUCN Red List categories and criteria (2010) was performed for all species for which taxonomic classification changed considerably. The geographic range of Ranitomeya was predicted using the ecological niche modeling algorithm and software (employing default settings) Maxent 3.3.3c of Phillips et al. (2006). Climatic data for model building were obtained through the WORLDCLIM website (http:// www.worldclim.org/). DNA extraction, amplification and sequencing. Tissue samples, collected from toe clippings from adult frogs and from tips of tadpole tails, were preserved in a buffer solution of 20% DMSO saturated with sodium chloride and EDTA. Genomic DNA was extracted from tissue samples using the Qiagen DNeasy Tissue Kit. The cytochrome-b gene (cytb), 12S ribosomal RNA (12S) and 16S ribosomal RNA (16S) regions were amplified using the following primer sets for a total of 1011 base pairs: cytb: CB1-L, CB2-H (Kocher et al. 1989); KSCYB1(A)-L, KSCYB1-H (Clough & Summers 2000); 12S: 12SA-L (Palumbi et al.1991), 12SK-H (Goebel et al.1999); 16S: LGL 286, LGL 381 (Palumbi et al. 1991). All loci used are from the mitochondrial genome. PCR amplifications and DNA sequencing protocols followed Roberts et al. (2006b). Samples were sequenced in both directions and aligned using Sequencher version 1.4.0 (ABI 1995). Sequences in both the nuclear and mitochondrial analyses were aligned using MUSCLE 3.6 (Edgar 2004). All sequences were translated to confirm proper reading frame and absence of stop codons. Mitochondrial phylogenetic analyses. Much of the sequence data used in this analysis were derived from previous studies (Summers et al. 1999; Clough & Summers 2000; Summers & Symula 2001; Santos et al. 2003; Symula et al. 2003; Darst & Cannatella 2004; Graham et al. 2004; Grant et al. 2006; Noonan & Wray 2006; Roberts et al. 2006a; Brown et al. 2008c, Santos et al. 2009). Newly determined sequences were deposited in Genbank under accession numbers JN651247–JN651274 and JN635752–JN635962 (see appendix I for details). Representative species of the genera Colostethus, Hyloxalus and Phyllobates (all sensu Grant et al. 2006) were used as outgroups for a phylogenetic analysis using Bayesian inference (Huelsenbeck & Ronquist 2001). The dataset was partitioned into codon-position specific sets of nucleotides (1st, 2nd, and 3rd positions for cytb, with a separate, single partition each for 12S and 16S). MrModeltest version 2.3 (Nylander 2004) was used to identify a substitution model, nucleotide frequencies and optimal priors for the gamma parameter and the proportion of invariant sites for each partition. Sequence data may better be explained by partitioning a dataset than by applying an average model across genes and codon positions, as indicated by higher model likelihood scores in partitioned analyses (Nylander et al. 2004). MrBayes version 3.1.1 (Huelsenbeck & Ronquist 2001) was run for 25 million generations using the following mixed models: 12S (nst = 6, I+Γ), 16S (nst = 6, I+ Γ), cytb-1st (nst = 1, equal), cytb-2nd (nst = 6, Γ), and cytb-3rd (nst = 6, Γ). All analyses were run in parallel on 8 CPUs, each chain on a separate processor, selected from available CPUs that are part of an 800+ cluster of computers at Duke University. Negative log-likelihood (-ln) scores were used to identify the burn-in phase at 12 million generations, after which all trees (n = 11,540) were summed. To ensure that the burn-in phase was properly defined in all runs, Bayesian parameter files were also analyzed in TRACER 1.3 (Rambout & Drummond 2006). The Bayesian analysis standard deviation of split frequencies reached < 0.01 to ensure consistency of the resulting tree topology.

TAXONOMIC REVISION OF RANITOMEYA

Nuclear phylogenetic analyses. All the sequence data used in this analysis were derived from Grant et al. (2006). A total of 2262 characters were used from five nuclear genes/markers: rhodopsin exon 1 (Rhod), recombination activating gene 1 (RAG1), histone H3 (H3), 28S rDNA (28S) and seventh in absentia (SIA). Representative species of the genera Hyloxalus and Phyllobates were used as outgroups for a phylogenetic analysis using Bayesian inference, as described above. The dataset was partitioned into codon-position specific sets of nucleotides (1st, 2nd, and 3rd positions for Rhod, RAG1, SIA and H3 with a separate, single partition for 28S). MrModeltest version 2.3 (Nylander 2004) was used to identify a substitution model, nucleotide frequencies and optimal priors for the gamma parameter and the proportion of invariant sites for each partition. MrBayes version 3.1.1 (Huelsenbeck & Ronquist 2001) was run for 15 million generations using the following mixed models: Rhod-1st (nst = 1, equal), Rhod-2nd (nst = 1, equal), Rhod- 3rd (nst = 6, propinv), RAG1-1st (nst = 1, equal), RAG1-2nd (nst = 1, equal), RAG1-3rd (nst = 6, Γ), SIA-1st (nst = 1, equal), SIA-2nd (nst = 2, Γ), SIA-3rd (nst = 6, equal), H3-1st (nst = 1, equal), H3-2nd (nst = 2, Γ), H3-3rd (nst = 6, equal), and 28S (nst = 1, I+Γ). Negative log-likelihood (-ln) scores were used to identify the burn-in phase at 4 million generations, after which all trees (n = 10,500) were summed. The Bayesian analysis standard deviation of split frequencies reached < 0.01 to ensure consistency of the resulting tree topology. We also used Maximum Parsimony in PAUP* (Swofford 2002) to estimate the phylogenetic relationships using the same dataset. All resulting trees (n = 820) were used to generate a consensus tree. Character state reconstructions. Character state reconstruction methods used to diagnose lineages was performed in PAUP* using ACCTRAN optimization (Swofford 2002). All unambiguous characters were mapped for the node of interest. Monophyly testing. To test the monophyly of groups we used Bayes Factor Analyses (BFA). This method differs from other traditional tests because it does not offer a criterion for absolute rejection of a null hypothesis, but instead it evaluates the phylogenetic evidence in favor of the null hypothesis. Bayes factor analyses were done in MrBayes by generating constraint trees of each test topology. The constraint trees each consisted of a tree containing at least, two hypothetical clades (both polytomies), one of which is the test group (e.g., a genus), while the other consists of all other taxa included in the analysis. This arrangement resulted in an absolute prior (= 100) for test topology. The predictive value of the constrained post burn-in harmonic mean likelihoods (H0) were then compared to the original, unconstrained post burn-in likelihoods (HA) using a Bayes factor comparison: 2 (Harmonic Mean Ln Likelihood HA − Harmonic Mean Ln Likelihood H0; Kass & Raftery 1995). Values above 10 were interpreted as strong evidence against H0. Values between 0 and 10 were interpreted as support for both hypotheses, and values less than 0 were interpreted as support against HA (Kass & Raftery 1995; Brandley et al. 2005; Suchard et al. 2005). Recent criticisms have challenged the use of harmonic means to calculate the marginal likelihood for model estimation; in particular, harmonic means can favor more parameter-rich models (Lartillot & Philippe 2006) and they can possess a systematic bias capable of significantly skewing Bayes factors (Calderhead & Girolami 2009). Given that the calculation of more robust estimates of the marginal likelihood (i.e., thermodynamic integration and path sampling) is statistically non-trivial, we chose to continue to use harmonic means, though we limit our interpretation of these results. Further, we compared our results to other published studies using other tests of monophyly. In many cases the estimated marginal likelihoods varied considerably for hypotheses regarding a particular clade: the hypothesis most similar to our unconstrained topology had an estimated marginal likelihood similar to the unconstrained tree and those with more dramatic changes were much larger. We evaluated the following 16 hypotheses (Table 2) using BFA: H1, Ranitomeya and Andinobates as reciprocally monophyletic clades using the nuclear dataset; H2, Ranitomeya (sensu this paper) monophyletic; H3, Andinobates monophyletic; H4, Andinobates sister to Excidobates; H5, Andinobates sister to the vanzolinii group; H6, Andinobates sister to a clade comprising the defleri, ventrimaculata and reticulata groups; H7, Ranitomeya biolat and R. lamasi (including R. sirensis) as reciprocally monophyletic clades; H8, R. biolat within R. lamasi (including R. sirensis); H9, R. sirensis and R. lamasi (including R. biolat) as reciprocally monophyletic clades; H10, R. sirensis within R. lamasi (including R. biolat); H11, R. ventrimaculata (sensu this paper) monophyletic; H12, R. amazonica (sensu this paper) monophyletic; H13, R. variabilis (containing only individuals from Cainarachi Valley) monophyletic; H14, R. variabilis and R. variabilis cf. (all spotted R. ventrimaculata) monophyletic; H15, R. amazonica individuals from French Guiana and all other R. amazonica individuals as reciprocally monophyletic sister clades; H16, R. flavovittata monophyletic.

BROWN ET AL.

TABLE 1A (continued on next page). Currently accepted taxa (based on Frost et al. 2011). History of alpha-taxonomic and species group (in parentheses) classification of taxa in the genera Ranitomeya and Andinobates. Dendrobates (minutus)* = by implication, "+" species which were not described at the time of classification, "*" species which were not acknowledged by the authors. Specific Epithet

Description (authors and date)

fantasticus

Boulenger, 1884 “1883”

reticulatus

Boulenger, 1884 “1883”

opisthomelas

Boulenger, 1899

minutus

Shreve, 1935

Dendrobates (minutus)

ventrimaculatus

Shreve, 1935

Dendrobates (minutus)* Fig. 14 j

Silverstone 1975 Dendrobates (minutus)* Fig. 14: I, h?, l, m Dendrobates (minutus)* Fig. 14: a, b, c Dendrobates (minutus)

Classification: Genus (group) Myers & Daly Myers 1982 1980 * Dendrobates (minutus) Dendrobates (minutus) Dendrobates (minutus) *

Dendrobates (quinquevittatus) Dendrobates (quinquevittatus) Dendrobates (minutus) Dendrobates (minutus) *

Myers 1987 Dendrobates (quinquevittatus) Dendrobates (quinquevittatus) Minyobates Minyobates *

igneus

Melin, 1941

Dendrobates (minutus)* pg. 34

altobueyensis

Silverstone, 1975

Dendrobates (minutus)

Minyobates

fulguritus

Silverstone, 1975

Dendrobates (minutus)

abditus

Myers & Daly, 1976

Dendrobates (minutus) Dendrobates (minutus) Dendrobates (minutus)

Dendrobates (minutus)* Fig. 14 e

Myers & Daly, 1976

Myers & Daly, 1980

Dendrobates (minutus) Dendrobates (minutus)

Dendrobates (minutus) Dendrobates (minutus) Dendrobates (minutus) Dendrobates (quinquevittatus) Dendrobates (minutus) Dendrobates (minutus)

imitator

Schulte, 1988

variabilis

sirensis

Zimmermann & Zimmermann, 1988 Aichinger, 1991

biolat

Morales, 1992

lamasi

Morales, 1992

virolinensis

amazonicus

Ruiz-Carranza & Ramírez-Pinilla, 1992 Schulte, 1999

duellmani

Schulte, 1999

flavovittatus

Schulte, 1999

intermedius

Schulte, 1999

rubrocephalus

Schulte, 1999

claudiae

Jungfer, Lötters & Jörgens, 2000

Dendrobates (minutus)*

Minyobates*

daleswansoni

Rueda-Almonacid, Rada, Sánchez-Pacheco, VelásquezÁlvarez & Quevedo-Gil, 2006 Rueda-Almonacid, Rada, Sánchez-Pacheco, VelásquezÁlvarez & Quevedo-Gil, 2006 Brown, Schulte & Summers, 2006

Dendrobates (minutus)* pg. 34

Dendrobates (minutus)* Fig. 14 k

Dendrobates (minutus)* Fig. 14: h?, g

Dendrobates (minutus)* Fig. 14 f

vanzolinii

Myers, 1982

viridis bombetes

dorisswansonae

uakarii tolimense benedicta summersi cyanovittata defleri

Bernal-Bautista, Luna-Mora, Gallego & Quevedo-Gil, 2007 Brown, Twomey, Pepper & Sanchez-Rodriguez, 2008 Brown, Twomey, Pepper & Sanchez-Rodriguez, 2008 Perez-Peña, Chavez, Twomey & Brown, 2010 Twomey & Brown, 2009

Minyobates Minyobates Dendrobates (quinquevittatus) Minyobates Minyobates

steyermarki

Perez-Peña, Chavez, Twomey & Brown, 2010 Rivero, 1971

Dendrobates (minutus)

Steindachner, 1864

Dendrobates (minutus)

Dendrobates (minutus) Dendrobates (quinquevittatus)

Minyobates

quinquevittatus

Dendrobates (minutus) Dendrobates (minutus)

yavaricola

TAXONOMIC REVISION OF RANITOMEYA

Dendrobates (quinquevittatus)

TABLE 1B (continued from previous page). Current

Classification: Genus (group) Specific Epithet

Caldwell & Myers 1990

fantasticus

Schulte 1999

Grant et al. 2006

Dendrobates (Group 1)

Ranitomeya (ventrimaculata) Ranitomeya (ventrimaculata) Ranitomeya (minuta) Ranitomeya (minuta) Ranitomeya (ventrimaculata) Ranitomeya (ventrimaculata) Ranitomeya (minuta) Ranitomeya (minuta) Ranitomeya (minuta) Ranitomeya (ventrimaculata) Ranitomeya (minuta) Ranitomeya (minuta) Ranitomeya (ventrimaculata) Ranitomeya (ventrimaculata) Ranitomeya (ventrimaculata) Ranitomeya (ventrimaculata) Ranitomeya (ventrimaculata) Ranitomeya (minuta) Ranitomeya (ventrimaculata) Ranitomeya (ventrimaculata) Ranitomeya (ventrimaculata) Ranitomeya (ventrimaculata) Ranitomeya (ventrimaculata) Ranitomeya (minuta) *

alphaTwomey & Brown 2008

Dendrobates (ventrimaculatus) Dendrobates (ventrimaculatus) Minyobates

Dendrobates (Group 6)

Minyobates

Dendrobates (ventrimaculatus) *

Dendrobates (Group 5)

altobueyensis

Minyobates

fulguritus

Minyobates

abditus

Minyobates

Dendrobates (ventrimaculatus) Minyobates

Dendrobates (Group 4)

bombetes

Minyobates

imitator

Dendrobates (Group 3)

variabilis

Dendrobates (ventrimaculatus)

Dendrobates (Group 1)

sirensis

Dendrobates (Group 6)

biolat

Dendrobates (Group 4)

lamasi

Dendrobates (Group 4)

virolinensis

Dendrobates (Group 6)

amazonicus

Dendrobates (Group 1)

duellmani

Dendrobates (Group 1)

flavovittatus

Dendrobates (Group 2)

intermedius

Dendrobates (Group 3)

rubrocephalus

Dendrobates (Group 6)

Minyobates*

daleswansoni

dorisswansonae

uakarii

Ranitomeya (reticulata)

tolimense

benedicta

summersi

Ranitomeya (reticulata)*

cyanovittata

defleri

yavaricola

Minyobates Dendrobates (quinquevittatus)

Dendrobates (Group 6) Dendrobates (Group 2)

Minyobates Adelphobates

reticulatus opisthomelas minutus ventrimaculatus igneus

vanzolinii viridis

claudiae

steyermarki quinquevittatus

14 路 Zootaxa 3083 漏 2011 Magnolia Press

Ranitomeya (reticulata) Ranitomeya (reticulata) * Ranitomeya (minuta) Ranitomeya (ventrimaculata) * * Ranitomeya (ventrimaculata) * Ranitomeya (vanzolinii) * * Ranitomeya (vanzolinii) Ranitomeya (ventrimaculata) Ranitomeya (vanzolinii) Ranitomeya (vanzolinii) Ranitomeya (vanzolinii) * Ranitomeya (ventrimaculata) Ranitomeya (reticulata) Ranitomeya (vanzolinii) Ranitomeya (vanzolinii) Ranitomeya Ranitomeya (minuta) *

This paper Ranitomeya (reticulata) Ranitomeya (reticulata) Andinobates (bombetes) Andinobates (minutus) Ranitomeya (ventrimaculata) Ranitomeya (reticulata) Andinobates (fulguritus) Andinobates (fulguritus) Andinobates (bombetes) Ranitomeya (vanzolinii) Andinobates (fulguritus) Andinobates (bombetes) Ranitomeya (vanzolinii) Ranitomeya (ventrimaculata) Ranitomeya (vanzolinii) Ranitomeya (vanzolinii) Ranitomeya (vanzolinii) Andinobates (bombetes) Ranitomeya (ventrimaculata) Ranitomeya (reticulata) Ranitomeya (vanzolinii) Ranitomeya (vanzolinii) Ranitomeya Andinobates (minutus) Andinobates (bombetes) Andinobates (bombetes) Ranitomeya (reticulata) Andinobates (bombetes) Ranitomeya (reticulata) Ranitomeya (reticulata) Ranitomeya (vanzolinii) Ranitomeya (reticulata) Ranitomeya (vanzolinii) Minyobates Adelphobates

taxonomic status fantastica reticulata opisthomelas minutus ventrimaculata reticulata altobueyensis fulguritus abditus vanzolinii viridis bombetes imitator variabilis sirensis sirensis sirensis virolinensis amazonica ventrimaculata flavovittata imitator nomen dubium claudiae daleswansoni dorisswansonae uakarii tolimensis benedicta summersi cyanovittata defleri yavaricola steyermarki quinquevittatus

BROWN ET AL.

Due to the aforementioned concerns regarding BFA, we additionally performed Shimodaira-Hasegawa tests for many of the 16 hypotheses (see Table 3), though given the differences between methods and the resulting tests, some hypotheses are not identical and some caution should be taken when comparing these results. These analyses were performed in PAUP* (Swofford 2002) and test topologies were generated in MacClade (Maddison & Maddison, 2001). Bioacoustic analysis. We gathered call data from 21 species of Andinobates and Ranitomeya. For four of these species, we relied solely on previously published data. For the remaining 17 species, we analyzed new calls in addition to previously published data. Of these 17 species, 12 are having their calls described herein for the first time. In total, we obtained calls from 66 individuals of 17 species. Calls were recorded by six different people. Recordings by J.L. Brown and E. Twomey used either a Sony TCM 5000 EV tape recorder or a Marantz PMD660 solid state digital recorder with a Sennheiser ME 66-K6 microphone. Recordings by D. Mejia-Vargas used a Sony TCM-5000 EV cassette recorder. E. Poelman made available recordings from Marty & Gaucher (1999) of French Guianan R. amazonica. Recordings by Adolfo Amézquita used a Marantz PMD660 solid state digital recorder with a Sennheiser ME 64 unidirectional microphone. Recordings by T. Ostrowski used a Sony Professional walkman (WM-D6C) and a Sennheiser MKH 416 P48 microphone, or an iPhone4 in combination with the software Record Pad 5.04. For details on calls originating from previous publications, see references provided in Table 5. The following bioacoustic variables were analyzed: dominant frequency (frequency at which peak amplitude occurs), note length (length in seconds of a note, where a note is as a discrete bundle of pulses), pulses per note (where each pulse is defined as a burst of energy, several of which together comprise a note), pulse rate (pulses per second, calculated as pulse count for a note divided by note length), internote interval (the amount of time from the end of one note to the beginning of the following note), and notes per minute (number of notes given during 1 minute, if the recording was less than one minute this was extrapolated from the available recording). In addition, we provided verbal descriptions of each call type. For the purposes of statistical analysis of call parameters between Ranitomeya amazonica and R. variabilis, the unit of replication is an individual frog, i.e., multiple measurements taken from the same individual were averaged to a single value. Calls were analyzed in Raven Pro 1.4 (Charif 2010).

Results Phylogenetics. Our mitochondrial phylogenetic topology is similar to recent studies (e.g. Grant et al. 2006 and Santos et al. 2009) and all major groups are well supported (Fig. 3, see discussion for comments on Minyobates ). Increased sampling reveals a more detailed phylogeography of many species throughout their known ranges. Within Ranitomeya, there exist four major monophyletic groups which we herein define as species groups: the variabilis, reticulata, defleri and vanzolinii groups (Fig. 3). In the variabilis group, there exist two reciprocally monophyletic clades (classified herein as R. amazonica and R. variabilis). Within R. variabilis (not to be confused with the variabilis group), there exists a deep divergence among populations predominantly from areas surrounding the upper Huallaga river (individuals 40-46) and all other populations (individuals 1-39). Similar biogeographic divergence is observed in other dendrobatids (e.g. between Ameerega pepperi and A. bassleri, and between R. fantastica and R. summersi), suggesting a barrier existed between upper Huallaga river populations and other populations in recent history. In the two cases mentioned, this likely resulted in allopatric speciation. For Ranitomeya variabilis, though minor morphological differences exist, these populations do not appear to be different species and it appears ancestors to these individuals have radiated northward, as far as Rio Marañón (individual 47) and east to the Serranía de Contamana (individual 42), mixing with local populations. In the reticulata group, there is deep divergence among the clade containing R. reticulata/R. ventrimaculata, the clade containing R. fantastica/R. summersi/R. benedicta and the R. uakarii clade. Within R. uakarii a northsouth break exists between Tournavista, PE/Porto Walter, BZ and Tamshiyacu-Tahuayo, PE. In Ranitomeya, the deepest genetic divergence between sister taxa exists between R. toraro sp. nov. and R. defleri. Both species comprise the defleri group, a group that is sister to the reticulata group which form a clade sister to the variabilis group. The vanzolinii group, a group sister to all other Ranitomeya groups, contains two divergent clades. One clade contains R. sirensis and the other contains the remaining five species in the group.

TAXONOMIC REVISION OF RANITOMEYA

...... continued on the next page

16 路 Zootaxa 3083 漏 2011 Magnolia Press

BROWN ET AL.

FIGURE 2. Illustrated guide to morphological terminology. A. Finger and hand morphology: i. Finger I (far left) < Finger II, thenar tubercle (= inner metacarpal tubercle) present (depicted by arrow), and greatly expanded finger discs in Fingers II-IV. Inset depicts Finger I and a thenar tubercle which is clearly visible. Note that in some Ranitomeya this is trait reduced and difficult to view (as in main picture) (Ranitomeya variabilis pictured, inset of R. benedicta). ii. Finger I ≈ Finger II, thenar tubercle absent. (Adelphobates quinquevittatus pictured) iii. Weakly expanded finger discs in Fingers II-IV (Excidobates captivus pictured). B. Stripes: i. Middorsal (follows vertebral column), dorsolateral (extends from eye to either upper thigh, as pictured, or to vent), ventrolateral (running from groin to axilla) and labial stripe (stripe that extends from shoulder around upper lip)(R. sirensis pictured). ii.. Oblique lateral stripe (extends from groin to eye, as in picture stripe is incomplete anteriorly). Unlabeled arrow depicts a dorsolateral stripe that does not reach thigh, a characteristic of certain species of Andinobates (type ‘A’ in Grant et al. 2006). (Andinobates claudiae pictured). C. Limb patterns: i. Distinct limb reticulation/spotting (characteristic of most species of Ranitomeya) (R. variabilis pictured). ii. Wavy stripes (not classified as distinct limb reticulation) (R. summersi pictured). iii. Patternless. Typical of most Andinobates species (R. sirensis pictured). D. Diagnostic head patterns: i. Large black “oval” on head (R. imitator pictured). ii. Large black “pentagon” or “five-point star” on head (R. summersi pictured). iii. Black band across head entirely covering eyes (known only in a single population of this species near the Pongo de Manseriche, Peru) (R. fantastica pictured). E. Nose spots. i. Two nose spots (R. imitator pictured). ii. Single nose spot. (R. variabilis pictured). iii. Frontward-turned “U” on the tip of snout. (R. toraro pictured). F. Geographical distribution. West: distribution within Andes, west of Andes, or in Central America. East: distribution east of Andes (including Guiana Shield) or in east-Andean versant. G. Dorsal patterns: i.“Y-shape”. Space between stripes create black pattern which forms a black Y on the back. (R. variabilis pictured). ii. Merging of the obliquelateral and dorsolateral stripes (R. variabilis pictured). iii. Broken dorsolateral stripes (R. flavovittata pictured). iv. Spotting (R. imitator pictured). H. Key ventral characters: i. Distinctive throat coloration and ventral reticulation (also shown in H-ii & H-iii) (R. reticulata pictured). ii. Belly patch (R. sirensis pictured). iii. Gular spots (single or paired dark spots at corner of mouth) (R. amazonica pictured). iv. Marbled pattern (not classified as reticulation) (Andinobates virolinensis pictured).

In Andinobates gen. nov, there appear to be three distinct clades (classified herein as species groups): bombetes, fulguritus and minutus. The minutus group is sister to both the fulguritus and bombetes groups. Within the bombetes group, there is little phylogenetic distinction between A. bombetes and A. virolinensis, suggesting they may be different morphs of the same species or have diverged recently. There is strong phylogenetic support for the placement of A. dorisswansonae as a member of the bombetes group, a species which, along with A. daleswansoni, was thought to deserve allocation to a new genus (Rueda-Almonacid et al. 2006) based on its unique foot morphology. Bioacoustics. Call data is summarized in Table 5. Within the genus Ranitomeya, we recognize three main call types: reticulata group calls, variabilis group calls, and vanzolinii group calls. Calls of the reticulata group and variabilis group sound similar, with the main difference that notes are given much more rapidly in the reticulata group. The vanzolinii group has a distinct call, consisting of a tonal trill. The call of R. defleri is similar to calls from the variabilis group, while the call of R. toraro remains unknown. Within Andinobates gen. nov, it seems that all calls are some permutation of a buzz or rattle. The two main call groups appear to correspond to the Andean and Chocoan clades. Both species in the minutus group (A. claudiae and A. minutus), as well as A. fulguritus (and likely A. viridis) have similar calls consisting of short buzzes with varying levels of tonality and relatively high energy. Species in the bombetes group have long, atonal calls sounding like a buzz or rattle. These calls are typically longer than 1 sec, whereas calls of the Chocoan Andinobates gen. nov have calls typically shorter than 1 sec. Within Ranitomeya, there is a strong correspondence of call type to species group, and three of the four species groups can be readily identified by call type. Differences between species within groups are more subtle; in some cases they can be diagnostic (see for example R. benedicta, Brown et al. 2008), whereas in other cases calls within a group appear to be relatively homogenous (for example, within the vanzolinii group). However, the overall trend is low call variation within, and high call variation between species groups. Within Andinobates gen. nov, the call groups appear to be less distinct, and in many cases our analyses are based on a single individual of a species, thus our understanding of intraspecific vs. interspecific variation is incomplete.

TAXONOMIC REVISION OF RANITOMEYA

Morphology. Gross morphology in this group is of limited use in determining relationships below genus level. While Andinobates gen. nov and Ranitomeya can be diagnosed from each other on the basis of several characters (e.g., vertebrae 2 & 3 fused in Andinobates gen. nov, limb reticulation present in Ranitomeya), differences between species groups within each genus are less obvious and better resolved using bioacoustic and phylogenetic data. Within Ranitomeya, the vanzolinii group is the most distinctive morphologically, but the only unambiguous synapomorphy, as far as we can tell, is egg color (white in the vanzolinii group vs. dark grey or black in all other Ranitomeya). Within species groups, color pattern can be of some use in differentiating species, but determining which characters are useful is often made apparent in light of a molecular phylogeny. This problem is compounded by the fact that Müllerian mimicry appears to be widespread throughout Ranitomeya and often distant relatives appear to be morphologically identical. Even in species which are not mimetic, such as R. sirensis, intraspecific variation in color pattern is so high that previous authors considered different color morphs to be distinct species. Within Andinobates gen. nov, the bombetes group can be diagnosed from the fulguritus + minutus group on the basis of a wide medial gap in the papillae on the posterior labium of larvae (gap present in bombetes group). We have tentatively defined the fulguritus group on their shared green dorsal coloration, however A. viridis and A. altobueyensis have not yet been sequenced, so the validity of this arrangement needs to be confirmed with molecular data.

Key to the ‘small’ aposematic poison frogs of South America This key is designed to be useful in the field with only basic, easily observable characters such as hand and foot morphology and color pattern. Because it is designed to be a field key, some knowledge of the geographic distribution is expected. Notes in brackets indicate additional information that can be useful for identification. 1a. 1b. 2a. 2b. 3a. 3b. 4a. 4b. 5a.

5b.

6a. 6b. 7a. 7b. 8a. 8b. 9a. 9b. 10a. 10b. 11a. 11b. 12a. 12b.

First finger approximately equal in length or longer than second finger (Fig. 2a-ii) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 First finger shorter than second finger (Fig. 2a-i) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Thenar tubercle absent (Fig. 2a-ii) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Thenar tubercle present (Fig. 2a-ii). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 White spots or dashes on dorsum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Adelphobates castaneoticus [bright orange flash marks onupper surfaces of limbs; distribution in state of Mato Grosso and Para, Brazil] White stripes on dorsum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Adelphobates quinquevittatus [limbs orange with black spots; distribution in States of Rondônia and Amazonas, Brazil and parts of Pando, Bolivia] SVL greater than 25 mm, entire body covered with white spots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Excidobates mysteriosus [distribution in Cordillera del Condor north of Bagua, Peru] SVL less than 18 mm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Flash marks absent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Minyobates steyermarki [dorsum and venter essentially solid red or reddish-brown, sometimes with faint dorsolateral stripes; known only from Cerro de Yapacana, Venezuela] Yellow flash marks present above axilla and groin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Excidobates captivus [dorsum with distinct reddish orange markings in the form of elongated dashes, dorsolateral stripes, or heavy flecking; venter with distinct yellow blotches; distribution in upper Río Marañón drainage in northern Peru and southeastern Ecuador] Distinct limb reticulation absent (Fig. 2c-ii and iii) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Distinct limb reticulation/spotting present (Fig. 2c-i) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Distribution in Andes or west of Andes (i.e., Chocoan Ecuador or Colombia), or in Central America (Fig. 2f, see “west”) . . . 8 Distribution east of Andes (including Guiana Shield) or in east-Andean versant (Fig. 2f, see “east”) . . . . . . . . . . . . . . . . . . . 20 Four toes on foot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Five toes on foot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Head entirely red, body dull gold or brown . . . . . . . . . . . . . . . . Andinobates daleswansoni [distribution in Caldas, Colombia] Body black with red vermiculations or spots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Andinobates dorisswansonae [some individuals with yellow spots on flanks; distribution in Tolima, Colombia] Dorsolateral stripes present (in some species these stripes are sometimes faint and/or incomplete) . . . . . . . . . . . . . . . . . . . . . 11 Dorsolateral stripes absent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Distinct, well-defined yellow or green dorsolateral and middorsal stripes (occasionally stripes can be orange or with a broken mid-dorsal stripe) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Dorsal stripes diffuse; reddish or orange. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Limbs pale brown, dorsal stripes tan to bright yellow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Andinobates claudiae [SVL less than 14 mm; distribution in Bocas del Toro Province of Panama] Limbs black with distinct marbling identical in color to dorsal coloration; body bicolored (typically green, yellowish or bronze). . . . . . . . . . . . . . . . . . . . . . Andinobates fulguritus [distribution in eastern Panama and lowlands of western Colombia]

BROWN ET AL.

13a. 13b. 14a. 14b. 15a. 15b. 16a. 16b. 17a. 17b. 17c. 18a. 18b. 19a. 19b. 20a. 20b. 21a. 21b. 22a.

22b. 22c. 23a. 23b. 24a. 24b. 24c. 25a. 25b. 25c. 26a. 26b. 27a. 27b. 28a. 28b. 28c. 28d. 29a. 29b. 30a. 30b. 31a. 31b.

Stripes bright red, broad, usually on anterior half of dorsum, pale blue markings often present on flanks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Andinobates bombetes [distribution in Valle del Cauca, Quindío and Risaralda, Colombia] Stripes run nearly length of entire dorsum, typically orange-red, distinct stripe present extending along flank . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Andinobates minutus [distribution in central and eastern Panama, extending into Chocoan Colombia] Dorsum uniformly colored, lacking markings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Dorsum with black markings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Entire body one color (except for axilla and groin) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Body two or more colors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Entire body emerald green . . . . . Andinobates viridis [distribution in western slope of Colombian Andes and southern Chocó] Entire body yellow or metallic gold . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Andinobates altobueyensis [known only from the summit of Alto del Buey, Chocó, Colombia] Dorsum solid red, venter brown, flanks often reticulated blue and black. . . . . . . . . . . . . . . . . . . . . . . Andinobates opisthomelas [distribution in Antioquia, Caldas, Córdoba, and Chocó, Colombia] Dorsum dull orange-red, distinct orange labial stripe and markings on upper surface of thigh and upper arms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Andinobates minutus [distribution in central and eastern Panama, extending into Chocoan Colombia] Dorsal coloration fades from a brightly colored head to a dull posterior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Head yellow, fading to brown . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Andinobates tolimensis [distribution in Tolima, Colombia] Head and most of body bright crimson red, fading to brown on rump, venter blue with black reticulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Andinobates virolinensis [distribution in Cudinamarca and Santander, Colombia] Spots very small, body color yellow or metallic gold. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Andinobates altobueyensis [known only from the summit of Alto del Buey, Chocó, Colombia] Spots elongated, large, body mostly bright green . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Andinobates fulguritus [distribution in eastern Panama and lowlands of western Colombia] Body and limbs one solid color . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Andinobates abditus [body and limbs bronzy brown or black, orange markings in axilla and groin, known only from Volcán Reventador, Ecuador] Body and limbs different colors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Limbs bicolored. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Limbs one solid color, lacking distinct markings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Head orange, large black “oval” on head (Fig. 2d-i), call consisting of a musical “trill”. . . . . . . . . . . . . . . Ranitomeya imitator [this morph is a Müllerian mimic of R. summersi, therefore, diagnosis is difficult without call data; occurs near the towns of Chazuta and Sauce, San Martín, Peru] Black band across head entirely covering eyes (Fig. 2d-iii) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ranitomeya fantastica [similar morphologically to R. summersi; this morph is known only from the Pongo de Manseriche, Loreto, Peru] Head orange, large black “pentagon” present on head (Fig. 2d-ii) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 White lines on posterior half of body . . . . . . . . . . . . . . Ranitomeya fantastica [morph known from vicinity of Tarapoto, Peru] Orange lines on posterior half of body Ranitomeya summersi [occurs near the towns of Chazuta and Sauce, San Martín, Peru] Dorsum red, large red blotch on belly, legs metallic turquoise-green. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ranitomeya sirensis [nominotypical morph, known from the Cordillera El Sira, Peru] Black dorsum with pale greenish-blue spots, legs solid bronze . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ranitomeya yavaricola [known from Rio Yavari region of Peru] Legs solid blue or black (some individuals with blue-black gradient), conspicuous red head . . . . . . . . . . Ranitomeya benedicta [this morph occurs along the Río Ucayali and parts of the Cordillera Azul in southern Loreto, Peru] Dorsum with regular, complete, longitudinal stripes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Dorsum with irregular, broken stripes and/or spots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 Dorsum with neither stripes, spots, nor dashes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 Stripes create black pattern which forms a distinct Y-shape on the back (Fig. 2g-i) (the colored middorsal stripe is typically incomplete, confined to anterior half of body) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Stripes essentially parallel, middorsal stripe complete (Fig. 2b-i) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 From the Iquitos region and has red or orange stripes, or from the Guiana Shield and has yellow stripes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ranitomeya amazonica Stripes yellow or green, distribution in Peru, Ecuador, or Colombia (east Andean versant and Amazonian lowlands) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ranitomeya variabilis Stripes blue on black background . . . . . Ranitomeya cyanovittata [known only from Sierra del Divisor region of eastern Peru] Stripes green on black background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ranitomeya sirensis [this morph has been found near the towns of Contamana, Loreto, Peru and Codo del Pozuzo, Huánuco, Peru] Stripes yellow on black background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Stripes red or orange on black background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Yellow patch on belly present . . . . . . . . . . . . . . . . . . . . . Ranitomeya sirensis [occurs widely throughout much of central Peru] No patch on belly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Distribution in lowlands near Tarapoto and Yurimaguas, Peru. . . . . . . . . . . . . . . . . . . . .Ranitomeya imitator [musical trill call] Distribution east of Río Ucayali in Peru, throughout much of western Brazil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Frontward-turned ‘U’ on tip of snout (Fig. 2e-iii) . . . . . . . . . . . . . . . .Ranitomeya toraro sp. nov. [stripes straight and parallel] Round black spot on tip of snout (Fig. 2e-ii), stripes often wavy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ranitomeya uakarii [yellow-striped morph known only from Río Los Amigos in southern Peru and Porto Walter in western Brazil]

TAXONOMIC REVISION OF RANITOMEYA

32a. 32b. 33a. 33b. 34a. 34b. 35a. 35b. 36a. 36b. 37a. 37b. 38a. 38b. 39a. 39b. 40a. 40b. 41a. 41b. 41c. 42a. 42b.

43a. 43b.

Yellow, orange, or red patch on belly present . . . Ranitomeya sirensis [known from the vicinity of Puerto Inca, Huánuco, Peru] No patch on belly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 Yellow oblique lateral stripe present . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ranitomeya uakarii [known from Río Tahuayo region of northern Peru and one locality in the Río Pachitea drainage in central Peru] No yellow oblique lateral stripe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 SVL greater than 17 mm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Ranitomeya ventrimaculata [known from northern Peru, near the Ecuadorian border and from southern Ecuador] SVL less than 17 mm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ranitomeya reticulata [striped individuals of this species are typically rare; occurs near Iquitos, Loreto, Peru] Dorsum with black spots or dashes on bright background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 Dorsum with yellow spots or dashes on black background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Dorsum mostly red, few black spots present. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ranitomeya reticulata [spotted adults are rare, but this pattern is fairly common in juveniles; distribution near Iquitos, Peru] Dorsum mostly yellow, green, or blue with black spots present. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 Belly patch present . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ranitomeya sirensis [this morph is known from highlands of the upper Rio Iscozacín and Río Cacazú drainages, Pasco, Peru] Belly patch absent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 One nose-spot (Fig. 2e-ii), insect-like buzz call . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ranitomeya variabilis [this morph is typically found in montane regions in the east Andean versant of Peru and Ecuador] Two nose-spots (Fig. 2e-i), musical trill call. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ranitomeya imitator [nominotypical morph, known from the Cainarachi valley near Tarapoto, Peru] Belly patch present; spots yellow, small and circular . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ranitomeya vanzolinii [some individuals have an irregular middorsal stripe; distribution east of Río Ucayali in central Peru and extreme western Brazil] Belly patch absent; spots yellow, irregular; dashes or broken stripes also present . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 Limb and ventral reticulation bright blue, distribution near Río Apaporis, Colombia . . . . . . . . . . . . . . . . . . .Ranitomeya defleri [call is an insect-like buzz] Limb and ventral reticulation grayish-white, sometimes pale sky blue, distributed near Iquitos, Peru . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Ranitomeya flavovittata [call is a musical trill] Bright coloration on dorsum is about equal in area to black, dorsal coloration is bright orange,electric blue reticulation on limbs, trill call . . . . Ranitomeya imitator [this morph is only known from San Gabriel de Varadero, west of Yurimaguas, Peru] Bright coloration on dorsum confined to the head . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Dorsum almost entirely brightly colored . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Head coloration brilliant red, red coloration not extending to upper arms . . . . . . . . . . . . . . . . . . . . . . . . . Ranitomeya benedicta [body and limbs black with bright blue reticulation, distribution east of Río Huallaga and west of Río Ucayali in north-central Peru] Head coloration orange to copper, extending onto upper arms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ranitomeya fantastica [body black with light blue or white reticulation; known from Río Paranapura west of Yurimaguas, Peru and Cainarachi valley, north of Tarapoto, Peru] Dorsum bright red, SVL less than 17 mm. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ranitomeya reticulata [this is the most common morph, having a solid red dorsum; distribution near Iquitos, Peru] Dorsum reddish orange, SVL greater than 17 mm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ranitomeya amazonica [this morph is only known from a few undisclosed localities in Loreto, Peru]

Systematics Andinobates Twomey, Brown, Amézquita & Mejía-Vargas, gen. nov. Account authors: E. Twomey, J.L. Brown, A. Amézquita, D. Mejía-Vargas Figs. 3–8 Tables 2–6 Type species. Dendrobates bombetes Myers & Daly 1980 Etymology. From the Spanish adjective Andino (of the Andes) + the Greek bates (a walker), referring to the distribution of this genus, being primarily in the northern Andes. Gender masculine. The second half of the name is common among dendrobatine poison frogs. Proposed sister group. Ranitomeya. Definition and diagnosis. Unambiguous synapomorphies include: 15 nuclear and 11 mitochondrial synapomorphies (based upon the dataset used in this study, Fig. 3, Tables 2 and 3); vertebrae 2 and 3 fused (known only in A. claudiae, A. minuta, A. opisthomelas, and A. virolinensis; trait also present in Oophaga). Other characteristics include: adult SVL less than 20 mm; adults typically with bright coloration, often with bright red, green, or yellow dorsal coloration (either uniform or as stripes); Andinobates abditus an exception, dorsum mostly black or brown,

BROWN ET AL.

...... continued on the next page

TAXONOMIC REVISION OF RANITOMEYA

Zootaxa 3083 漏 2011 Magnolia Press 路

...... continued on the next page

22 路 Zootaxa 3083 漏 2011 Magnolia Press

BROWN ET AL.

...... continued on the next page

TAXONOMIC REVISION OF RANITOMEYA

Zootaxa 3083 漏 2011 Magnolia Press 路

FIGURE 3. A consensus Bayesian phylogeny based on 1011 base pairs of aligned mitochondrial DNA sequences of the 12S (12s rRNA), 16S (16s rRNA) and cytb (cytochrome-b gene) regions. Thickened branches represent nodes with posterior probabilities 90 and greater, other values are shown on nodes. Taxon labels depict current specific epithet, number in tree, the epithet being used prior to this revision (contained in parentheses), and the collection locality. A. Top segment. B. Middle segment. C. Bottom segment of phylogeny.

orange spots in axilla and groin. When present, complete dorsolateral stripe ending before thigh (type ‘A’ in Grant et al. 2006); ventrolateral stripe absent; when present, oblique lateral stripe incomplete. Limbs usually dark with or without pale longitudinal stripes, lacking discrete pale reticulation as in most species of Ranitomeya. Ventral coloration variable, usually with distinct, bright markings. Colored throat patch absent. Head narrower than body; teeth absent; vocal slits present in males. First finger distinctly shorter than second; finger discs II and III weakly to moderately expanded; toe disc III and IV weakly expanded; toe V unexpanded; toe webbing absent; median lingual process absent; tadpoles with LTRF 2(2)/3 (sometimes with gap in first posterior row, Table 4); larval oral disc emarginated; larvae with complete papillae on the posterior labium (so far known in A. minuta and A. claudiae) or wide medial gap in papillae on posterior labium (so far known in A. abditus, A. bombetes, A. opisthomelas, A. virolinensis and A. tolimensis). Distribution. This genus occurs within the rainforests of Colombia (Departments: Antioquia, Chocó, Santander, Cundinamarca, Caldas, Cauca, Córdoba, Tolima, Valle del Cauca, Quindío, Risaralda), Ecuador (Provinces: Napo) and Panama (Provinces: Bocas del Toro, Colón, Coclé, Kuna Yala, Veraguas).

BROWN ET AL.

FIGURE 4. Putative species tree for Andinobates, Excidobates, and Ranitomeya. Placement of species where molecular data were lacking (A. altobueyensis, A. viridis, A. abditus, A. daleswansoni and R. opisthomelas) was based on morphology. Andinobates altobueyensis and A. viridis were placed as sister taxa due to the absence of dark pigmentation on dorsal body and limbs and overall similar dorsal coloration and patterning. These species were placed as sister to A. fulguritus (sequenced) on the basis of similar dorsal coloration (bright green to greenish-yellow). Andinobates opisthomelas was placed in the bombetes group in a polytomy with A. bombetes and A. virolinensis (both sequenced) due to their similar advertisement calls and morphology, particularly their red dorsal pattern and marbled venter. Andinobates daleswansoni was placed as sister to A. dorisswansonae due to the absence of a well-defined first toe in both species. Andinobates abditus was placed in the bombetes group based on a larval synapomorphy which appears to be diagnostic of that group (wide medial gap in the papillae on the posterior labium). However, A. abditus was placed as the sister species to all other members of the bombetes group due to the absence of bright dorsal coloration and isolated geographic distribution. Andinobates abditus is currently the only species of its genus known to occur in the east-Andean versant, thus its placement remains speculative until molecular data become available. Photo credits: Thomas Ostrowski, Karl-Heinz Jungfer, Victor Luna-Mora, Giovanni Chaves-Portilla.

TAXONOMIC REVISION OF RANITOMEYA

Zootaxa 3083 漏 2011 Magnolia Press 路

TABLE 2. Bayes Factor Analyses. Bayes factor values above 10 were interpreted as strong evidence against H0. Values between 0 and 10 were interpreted as support for both hypotheses, and values less than 0 were interpreted as support against HA. HA= an unconstrained phylogeny. Often when Bayes factor values were between 0-10, Ho consensus topologies where similar to HA. Lower values often resulted in identical consensus topologies (differing only by branch lengths). * Includes R. sirensis sensu Aichinger 1991 +includes R. biolat sensu Morales 1992

Hypothesis (H0)

LnL: unconstrained

LnL: constrained

LnL: difference

Bayes factor [2ln(B10)]

Nuclear Dataset -5760.52

-5759.44

-1.08

-2.16

H2: Ranitomeya monophyletic (sensu this paper)

-13781.76

-13786.02

4.26

8.52

H3: Andinobates monophyletic

-13781.76

-13786.74

4.98

9.96

H4: Andinobates within Excidobates

-13781.76

-13839.19

57.43

114.86

H5: Andinobates sister to the vanzolinii group

-13781.76

-13797.16

15.40

30.80

H6: Andinobates sister to a clade comprising the defleri,

-13781.76

-13796.73

14.98

29.95

-13781.76

-13797.36

15.60

31.19

-13781.76

-13781.91

0.15

0.30

-13781.76

-13812.23

30.47

60.94

-13781.76

-13781.91

0.15

0.30

H11: R. ventrimaculata monophyletic (sensu this paper)

-13781.76

-13785.22

3.46

6.91

H12: R. amazonica monophyletic (sensu this paper)

-13781.76

-13784.14

2.76

4.76

H13: R. variabilis sensu Zimmermann & Zimmermann 1988

-13781.76

-13780.06

1.70

3.41

-13781.76

-13831.16

49.40

98.80

-13781.76

-13781.41

-0.35

-0.71

-13781.76

-13799.24

17.48

34.97

H1: Ranitomeya and Andinobates reciprocally monophyletic sister clades Mitochondrial Dataset

variabilis and reticulata groups

H7: R. biolat sensu Morales 1992 and R. lamasi* sensu Morales 1992 as reciprocally monophyletic sister clades

H8: R. biolat sensu Morales 1992 within R. lamasi* sensu Morales 1992

H9: R. sirensis sensu Aichinger 1991 sister to R. lamasi+ sensu Morales 1992

H10: R. sirensis sensu Aichinger 1991 within R. lamasi+ sensu Morales 1992

monophyletic (only individuals from Cainarachi Valley, Peru)

H14: All spotted and striped R. variabilis as reciprocally monophyletic sister clades

H15: R. amazonica individuals from French Guiana and all other R. amazonica individuals as reciprocally monophyletic sister clades

H16: R. flavovittata monophyletic

26 路 Zootaxa 3083 漏 2011 Magnolia Press

BROWN ET AL.

TABLE 3. Shimodaira-Hasegawa tests. * Includes R. sirensis sensu Aichinger. +includes R. biolat sensu Morales & excludes a R. amazonica individual 28 that falls as the base of this group. $ R. fantastica (including the two individuals from lower Huallaga) and R. summersi as sister taxa, which is sister to R. benedicta. xR. summersi sister to R. fantastica individuals that are finely reticulated in orange (i.e., individuals from the lower Huallaga valley) which are then sister to R. benedicta. This group forms a sister clade to all other R. fantastica.-parsimony tree length difference of 1. = parsimony tree length difference of 13. Unconstrained topology is from the Bayesian analysis (Fig. 3).

Hypothesis (H0)

LnL: unconstrained

LnL: constrained

LnL: difference

P value

Mitochondrial Dataset: this paper

H4: Andinobates within Excidobates

17048.31

17064.13

15.82

0.049

H5: Andinobates sister to the vanzolinii group

17048.31

17099.80

51.49

<0.001

H6: Andinobates sister to a clade comprising the defleri,

17048.31

17100.77

52.46

<0.001

17048.31

17174.21

125.90

<0.001

17048.31

17062.23

13.87

0.043

17048.31

17219.48

171.17

<0.001

17048.31

17379.14

330.83

<0.001

17048.31

17040.05

21.75

0.026

17048.31

17118.58

70.27

<0.001

17048.31

14048.87

1.85

0.491-

3071.48

3025.64

45.84

0.012=

variabilis and reticulata groups

H7: R. biolat sensu Morales 1992 and R. lamasi sensu Morales 1992* as reciprocally monophyletic sister clades

H9: R. sirensis sensu Aichinger 1991 and R. lamasi sensu Morales 1992+ as reciprocally monophyletic sister clades

H13: R. variabilis (only individuals from Cainarachi Valley, Peru) reciprocally monophyletic to all other striped R. variabilis

H14: All spotted R. variabilis reciprocally monophyletic to all other (striped) R. variabilis

H15: R. amazonica individuals from French Guiana and all other R. amazonica individuals as reciprocally monophyletic sister clades

H16: R. flavovittata and R. vanzolinii as reciprocally monophyletic sister clades H17: R. fantastica complex sensu Brown et al. 2008$ Mitochondrial Dataset from Brown et al. 2008 H18: R. fantastica complex sensu this paperx

TAXONOMIC REVISION OF RANITOMEYA

Zootaxa 3083 漏 2011 Magnolia Press 路

TABLE 4A (continued on next page). Basic Tadpole Morphology. Genus Andinobates

Species abditus

Body color gray

Tooth Row Formula 2(2)/3(1)

Andinobates

altobueyensis

gray

Andinobates

bombetes

gray

2(2)/3

Andinobates

claudiae

gray

Andinobates

daleswansoni

Marginal Papillae incomplete: absent dorsally and wide ventral gap *

Oral disc emarginate emarginate

2(2)/3(1)

incomplete: absent dorsally and wide ventral gap incomplete: wide dorsal gap

emarginate emarginate

gray

Andinobates

dorisswansonae

gray

Andinobates

fulguritus

gray

emarginate

Andinobates

minutus

gray

2(2)/3

incomplete: wide dorsal gap

emarginate

Andinobates

opisthomelas

gray

2(2)/3(1)

emarginate

Andinobates

tolimensis

gray

2(2)/3 (1)

Andinobates

viridis

incomplete: absent dorsally and wide ventral gap incomplete: absent dorsally and wide ventral gap *

Andinobates

virolinensis

dark brown

2(2)/3

Ranitomeya

amazonica

gray

2(2)/3 (1)

incomplete: wide dorsal and ventral gap incomplete: wide dorsal gap

emarginate * emarginate emarginate

Ranitomeya

benedicta

gray

2(2)/3

incomplete: wide dorsal gap

emarginate

Ranitomeya

cyanovittata

Ranitomeya

defleri

gray

2(2)/3(1)

incomplete: wide dorsal gap

emarginate

Ranitomeya

fantastica

gray

2(2)/3

incomplete: wide dorsal gap

emarginate

Ranitomeya

flavovittata

whitish

2(2)/3(1)

incomplete: wide dorsal gap

emarginate

Ranitomeya

imitator

whitish

2(2)/3(1)

incomplete: wide dorsal gap

emarginate

Ranitomeya

reticulata

gray

2(2)/3(1)

incomplete: wide dorsal gap

emarginate

Ranitomeya

sirensis

whitish

2(2)/3(1)

incomplete: wide dorsal gap

emarginate

Ranitomeya

summersi

gray

2(2)/3

incomplete: wide dorsal gap

emarginate

Ranitomeya

toraro

gray

2(2)/2(1)

incomplete: wide dorsal gap

emarginate

Ranitomeya

uakarii

gray

2(2)/3(1)

incomplete: wide dorsal gap

emarginate

Ranitomeya

vanzolinii

gray to dark gray

2(2)/3 (1)

incomplete: wide dorsal gap

emarginate

Ranitomeya

variabilis

gray

2(2)/3 (1)

incomplete: wide dorsal gap

emarginate

Ranitomeya

ventrimaculata

gray

Ranitomeya

yavaricola

whitish

2(2)/3 (1)

incomplete: wide dorsal gap

emarginate

28 路 Zootaxa 3083 漏 2011 Magnolia Press

BROWN ET AL.

TABLE 4B (continued from previous page). Basic Tadpole Morphology. Genus

SSpeciesecies

Nursery Pool Type

Larvaee Diet

CitatioCitation

Andinobates

abditus

large phytotelmata

Myers & Daly 1976

Andinobates

altobueyensis

large phytotelmata: axils of large aroids. * Species ID uncertain, possibly minutus

Silverstone 1975

Andinobates

bombetes

phytotelmata: bromeliads

Myers & Daly 1980

Andinobates

claudiae

phytotelmata: i.e., bamboo, “water filled * leaf axils”

Jungfer et al. 2000; Unpub. data Thomas Ostrowski

Andinobates

daleswansoni

phytotelmata: epiphytic bromeliads

Rueda-Almonacid et al. 2006

Andinobates

dorisswansonae

phytotelmata: epiphytic bromeliads

Rueda-Almonacid et al. 2006

Andinobates

fulguritus

phytotelmata: epiphytic bromeliads

Silverstone 1975

Andinobates

minutus

phytotelmata: epiphytic bromeliads

Silverstone 1975

Andinobates

opisthomelas

large phytotelmata: bromeliads

Silverstone 1975

Andinobates

tolimensis

Bernal-Bautista et al. 2007

Andinobates

viridis

Andinobates

virolinensis

large phytotelmata: large Tillandisa spp. * and Guzmania spp.

Ranitomeya

amazonica

phytotelmata: typically bromeliads (e.g., detritivorous, Pitcairnia geykessi, Aechmea aquilega, oophagous Catopsis berteroniana)

Ranitomeya

benedicta

phytotelmata: bromeliads

detritivorous, predaceous

Brown et al. 2008c

Ranitomeya

cyanovittata

Ranitomeya

defleri

medium size phytotelmata: bromeliads

detritivorous, likely predaceous

Twomey & Brown 2009

Ranitomeya

fantastica

large phytotelmata: Aechmea bromeliads and tree holes

detritivorous, predaceous

Brown et al. 2008c

Ranitomeya

flavovittata

medium phytotelmata: bromeliads

detritivorous, predaceous

this paper

Ranitomeya

imitator

small to medium phytotelmata: i.e., Heli- detritivorous, conia, Dieffenbachia and small tree holes oophagous

Ranitomeya

reticulata

small phytotelmata: i.e., bromeliads and detritivorous, predaceous seed capsules

Werner et al. 2010, J.L. Brown, unpub. data

Ranitomeya

sirensis

Medium phytotelmata: i.e., bromeliads, Xanthosoma, bamboo internodes

detritivorous, predaceous

von May et al. 2008b; J.L. Brown unpub. data

Ranitomeya

summersi

small-medium phytotelmata: i.e., Dieffenbachia

detritivorous, predaceous

Brown et al. 2008c

Ranitomeya

toraro

Phytotelmata of bromeliads i.e., detritivorous, predaceous Aechmea cf. bromeliifolia) and palms (i.e., Phenakospermum guyanense)

this paper

Ranitomeya

uakarii

Medium phytotelmata: i.e., Guzmania detritivorous, predaceous bromeliads

this paper

Ranitomeya

vanzolinii

small phytotelmata: i.e., tree holes, detritivorous, within woody vines oophagous

Ranitomeya

variabilis

Medium detritivorous, predaceous, this paper phytotelmata: Aechmea and Guzmania oophagous (variable) bromeliads, Dieffenbachia

Ranitomeya

ventrimaculata

Ranitomeya

yavaricola

small phytotelmata: i.e., bromeliads

detritivorous, predaceous

Perez-Peña et al. 2010

TAXONOMIC REVISION OF RANITOMEYA

Ruiz-Carranza & RamírezPinilla, 1992 predaceous, this paper

predaceous, this paper

predaceous, Caldwell & de Oliveira 1999

Vocalizations. Two main call groups within Andinobates correspond to the Andean and Chocoan clades (Fig. 8, Table 5). Species in the former (A. bombetes, A. opisthomelas, A. daleswansoni, A. dorisswansonae, A. tolimensis, A. virolinensis) produce calls that can be described as an extended rattle or buzz. These calls are typically over 1 sec in length and are relatively atonal with low energy. Chocoan species (A. claudiae, A. fulguritus, A. minutus) produce high-energy, short calls (less than 1 sec.) that are more reminiscent of calls produced by Ranitomeya species of the variabilis group. Data on A. abditus, A. viridis, A. virolinensis and A. altobueyensis are lacking. Species included (12). Dendrobates abditus Myers & Daly 1976; Dendrobates altobueyensis Silverstone 1975; Dendrobates bombetes Myers & Daly 1980; Dendrobates claudiae Jungfer, Lötters & Jörgens 2000; Dendrobates daleswansoni Rueda-Almonacid, Rada, Sánchez-Pacheco, Velásquez-Álvarez & Quevedo-Gil 2006; Dendrobates dorisswansonae Rueda-Almonacid, Rada, Sánchez-Pacheco, Velásquez-Álvarez & Quevedo-Gil 2006; Dendrobates fulguritus Silverstone 1975; Dendrobates minutus Shreve 1935; Dendrobates opisthomelas Boulenger 1899; Ranitomeya tolimensis Bernal-Bautista, Luna-Mora, Gallego & Quevedo-Gil 2007; Dendrobates viridis Myers & Daly 1976; Dendrobates virolinensis Ruiz-Carranza & Ramírez-Pinilla 1992. Remarks. Our definition of Andinobates is essentially equal to Myers’ (1987) definition of Minyobates minus steyermarki. When molecular data became available for Minyobates steyermarki (Vences et al. 2003; Roberts et al. 2006a), it was evident that this species was not affiliated with the other members of Minyobates sensu Myers (1987), but rather was more closely related to the Oophaga + Dendrobates + Adelphobates clade (see discussion for more information). The ancestor to Andinobates apparently diverged during the mid-Miocene from Excidobates about 17 mya and from Ranitomeya sensu this paper about 14 mya (Santos et al. 2009).

Species groups contained within Andinobates Andinobates minutus species group Figs. 3–5 (a–m), 8 Tables 4–6 A monophyletic assemblage of two divergent species: Andinobates minutus (Shreve 1935) comb. nov. and A. claudiae (Jungfer, Lötters & Jörgens 2000) comb. nov. Definition and diagnosis. SVL 13–16 mm; dark brown or black dorsum; largely complete light colored dorsolateral stripes not extending to thigh and oblique lateral stripes present, typically incomplete and not extending to eye; large spots near body on upper surface of legs and forearms; limbs and venter black with pale marbling on venter; LTRF either 2(2)/3(1) (A. claudiae) or 2(2)/3 (A. minutus); larvae with complete papillae on posterior labium and wide gap in papillae on anterior labium; oral disc emarginated (Table 4); larvae gray; eggs dark (Table 6). Vocalizations short, tonal buzz-notes, notes less than 1 sec in length, repeated at 3–17 notes per minute (Fig. 8, Table 5).

Andinobates fulguritus species group Figs. 3–5 (n–v), 6 (a–d), 8 Tables 2–4 An assemblage of three species: Andinobates altobueyensis (Silverstone 1975) comb. nov., A. fulguritus (Silverstone 1975) comb. nov. and A. viridis (Myers & Daly 1976) comb. nov. Definition and diagnosis. SVL 14–17 mm; body uniform yellow to dark green dorsally; A. fulguritus with black dorsolateral stripes and black oblique lateral stripes; A. altobueyensis lacking dorsal stripes but with fine black irregular spots on dorsum; A. viridis green with no black markings on dorsum; venter spotted or marbled in gray. Limbs identical to dorsal pattern and color. Vocalization data only available for A. fulguritus (Fig. 8, Table 5). Call consisting of short buzzes, notes 0.19–0.34 sec in length, repeated at 11–16 notes per minute. Call of A. viridis sounds similar to that of A. fulguritus, although recordings are lacking (D. MejíaVargas unpub. data, Table 5).

BROWN ET AL.

FIGURE 5. Andinobates Plate 1. minutus group: A–G: Andinobates claudiae and habitat (all from Bocas del Toro, Panama. Photos T. Ostrowski); A & B: Buena Esperanza; C–F: Isla Colon; G: Cerro Brujo; H: tadpole in phytotelm; I: habitat in Bocas del Toro, Panama. J–M: Andinobates minutus (all from Colombia. Photos DMV unless noted): J & K: Buenaventura, Valle del Cauca; L: Quibdó, Chocó; M: Baudó, Chocó (photo J. Mejía-Vargas). fulguritus group: N–V: Andinobates fulguritus (all from Colombia, photos DMV unless noted): N: Baudó, Chocó (photo J. Mejía-Vargas); O: Playa de Oro, Chocó (type locality); P–R: Uraba, Chocó. S–V: Anchicayá, Valle del Cauca. (nΦ = number of individual in phylogeny, Ω = population sampled in phylogeny). TAXONOMIC REVISION OF RANITOMEYA

FIGURE 6. Andinobates Plate 2. fulguritus group: A–D: Andinobates fulguritus (all from Risaralda, Colombia. photos DMV). bombetes group: E: Andinobates abditus type locality (photo W.E. Duellman). F–I: Andinobates daleswansoni and habitat (photos J. Mejía-Vargas); F & G: from type locality; H: type locality habitat; I: overview of habitat – human encroachment continues to threaten the habitat of this species. J–M: Andinobates dorisswansonae from Tolima, Colombia (photos DMV and T. Ostrowski). N–P: Andinobates tolimense from Tolima, Colombia (photos V. Mora-Luna). Q–R: Andinobates sp. aff. tolimense from Supatá, Colombia (photos G. Chaves-Portilla and T. Ostrowski).

BROWN ET AL.

FIGURE 7. Andinobates Plate 3. bombetes group: A–B: Andinobates sp. aff. tolimense; A: from Supatá, Colombia (Photo credits: G. Chaves-Portilla); B: Tolima, Colombia (Photo credits: Fundacion ProAves-Alonso Quevedo). C–H: Andinobates bombetes, all from Colombia (photo credits: DMV and T. Ostrowski); C: Quebrada a la Chapa, Boyacá; D-E: Valle del Cauca, T. Ostrowski; F: Yotoco, Valle del Cauca; G: Lowland habitat of A. bombetes near Quebrada a la Chapa; H: Yotoco, Valle del Cauca. I: Andinobates sp. aff. bombetes undisclosed locality in Colombia (photo credits: Dennis Nilsson and DMV). J–N: Andinobates virolinensis; Charalá, Virolin, Colombia . O–Q: Andinobates opisthomelas. O: Guatape, Antioquia, Colombia (photo credits: Dennis Nilsson and DMV); P-Q: Carmen de Atrato, Choco, Colombia.

TAXONOMIC REVISION OF RANITOMEYA

Table 5A (continued on following page). Summary of acoustic data in Ranitomeya and Andinobates. Species Andinobates altobueyensis Andinobates claudiae Andinobates fulguritus Andinobates viridis Andinobates minutus Andinobates abditus Andinobates bombetes Andinobates opisthomelas Andinobates daleswansoni Andinobates dorisswansonae Andinobates virolinensis Andinobates tolimensis Ranitomeya amazonica Ranitomeya variabilis Ranitomeya defleri Ranitomeya toraro Ranitomeya benedicta Ranitomeya fantastica Ranitomeya reticulata Ranitomeya summersi Ranitomeya uakarii Ranitomeya ventrimaculata Ranitomeya cyanovittata Ranitomeya flavovittata Ranitomeya imitator Ranitomeya sirensis Ranitomeya vanzolinii Ranitomeya yavaricola

Description unknown buzz short buzz similar to fulguritus tonal buzz unknown long buzz long buzz very similar to dorisswansonae long buzz long buzz buzz buzz buzz buzz unknown rapid buzz rapid buzz rapid buzz rapid buzz rapid buzz rapid buzz unknown trill trill trill trill trill

Dominant frequency

Note length

Pulses per note

5773–6079 Hz 4834–5161 Hz

0.85–1.03 sec 0.19–0.34 sec

55–65 28–52

5400–6400 Hz

0.2–1.1 sec

20–71

4000–4800 Hz 5000–5500 Hz

0.9–1.7 sec 1.6–2.3 sec

160 (estimate) 215–224

4100–4200 Hz

1.4–1.6 sec

72–78

4730–5220 Hz 4200–5800 Hz 4386–5624 Hz 5319–5415 Hz

0.84–0.99 sec 0.16–0.36 sec 0.14–0.44 sec 0.41–0.62 sec

n/a 20–49 19–116 40–61

3190–4240 Hz 2950–3790 Hz 4140–4480 Hz 2760–3220 Hz 3790–4130 Hz 4190–4400 Hz

0.10–0.17 sec 0.18–0.32 sec 0.18–0.29 sec 0.38–0.50 sec 0.26–0.29 sec 0.32–0.38 sec

n/a 10–13 pulses 48–94 pulses 14–16 14–16 pulses 58–63

5928–5950 Hz 4710–5660 Hz 5010–5690 Hz 5350–5440 Hz 5400–6000 Hz

0.81–1.06 sec 0.44–1.07 sec 0.88–2.2 sec 0.57–0.64 sec 0.63–0.88 sec

24–31 pulses 16–32 pulses 21–55 pulses 16–17 pulses 20–27 pulses

BROWN ET AL.

Table 5B (continued from previous page). Summary of acoustic data in Ranitomeya and Andinobates.

Species

Pulse rate

Inter-note interval

Notes per minute (extrapolated)

Source

Andinobates altobueyensis Andinobates claudiae

62–64 pulses/sec

18–24 sec

3 notes

Andinobates fulguritus

130–157 pulses/sec

3.1–5.9 sec

11–16 notes

Andinobates minutus

63–125 pulses/sec

1.2–3.2 sec

17 notes

Andinobates abditus Andinobates bombetes

113–134 pulses/sec

8.8 sec

5.76

Andinobates opisthomelas

92–95 pulses/sec

6–9 sec

6 notes

Andinobates dorisswansonae Andinobates virolinensis

47–52 pulses/sec

6.5 sec

7.5

Andinobates tolimensis

n/a

16.5 (estimate)

Ranitomeya amazonica Ranitomeya variabilis Ranitomeya defleri Ranitomeya toraro Ranitomeya benedicta Ranitomeya fantastica Ranitomeya reticulata Ranitomeya summersi Ranitomeya uakarii

85–138 pulses/sec 106–297 pulses/sec 94–104 pulses/sec

2.5 sec (estimate) 0.75–3.3 sec 0.6–2.0 sec 0.8–2.1 sec

Myers & Daly 1980; this study (calls courtesy D. Mejía-Vargas) Myers & Daly 1980; this study (calls courtesy A. Amézquita) this study (unpub data, D. Mejía-Vargas) this study (unpub data, D. Mejía-Vargas) this study (unpub data, D. Mejía-Vargas) Bernal et al. 2007

24–70 notes 34–62 notes 27 notes

this study this study Twomey & Brown 2009

n/a 41–57 pulses/sec 270–382 pulses/sec 39–40 pulses/sec 50–58 pulses/sec

0.13–0.18 sec 0.17–0.34 sec 0.12–0.15 sec 0.10–0.19 sec 0.10–0.11 sec

200 notes 107–146 notes 145–172 notes 104 notes 168 notes

Ranitomeya ventrimaculata

166–181 pulses/sec

0.12–0.13 sec

124 notes

Brown et al. 2008c this study this study Brown et al. 2008c Brown et al. 2006; this study this study (call courtesy Thomas Ostrowski)

Ranitomeya cyanovittata Ranitomeya flavovittata Ranitomeya imitator Ranitomeya sirensis Ranitomeya vanzolinii Ranitomeya yavaricola

29–30 pulses/sec 29–38 pulses/sec 24–30 pulses/sec 26–28 pulses/sec 31–32 pulses/sec

22–46 sec 4–20 sec 2–10 sec approx. 14 sec 7–37 sec

1.7 notes 7–11 notes 6–14 notes approx. 4 notes 2–7 notes

Andinobates viridis

Andinobates daleswansoni

TAXONOMIC REVISION OF RANITOMEYA

this study (unpub data, D. Mejía-Vargas) this study (calls courtesy T. Ostrowski and D. Mejía-Vargas) this study (unpub data, D. Mejía-Vargas) Myers & Daly 1976; this study (calls courtesy D. Mejía-Vargas)

this study this study this study this study Perez-Peña et al. 2010

Andinobates bombetes group Figs. 3, 4, 6 (e–r), 7 (a–q), 8 Tables 4–6 An assemblage of seven species: Andinobates abditus (Myers & Daly 1976) comb. nov., A. bombetes (Myers & Daly 1980) comb. nov., A. daleswansoni (Rueda-Almonacid, Rada, Sánchez-Pacheco, Velásquez-Álvarez & Quevedo-Gil 2006) comb. nov., A. dorisswansonae (Rueda-Almonacid, Rada, Sánchez-Pacheco, Velásquez-Álvarez & Quevedo-Gil 2006) comb. nov., A. opisthomelas (Boulenger 1899) comb. nov., A. tolimensis (Bernal-Bautista, Luna-Mora, Gallego & Quevedo-Gil 2007) comb. nov. and A. virolinensis (Ruiz-Carranza & Ramírez-Pinilla 1992) comb. nov.

FIGURE 8. Advertisement calls for species of Andinobates. A. Andinobates bombetes from Bosque Yotoco, Valle del Cauca, Colombia (type locality), recorded at 18-20° C; B. Andinobates claudiae from Isla Colón, Panama, recorded at 25° C (call courtesy Thomas Ostrowski); C. Andinobates fulguritus from Itauri, Colombia, unknown temperature; D. Andinobates fulguritus from Kuna Yala, Panama, recorded in captivity at 24° C (call courtesy T. Ostrowski); E. Andinobates dorisswansonae from “El Estadero”, Caldas, Colombia (type locality), recorded at 19-20° C; F. Andinobates minutus, unknown locality or temperature; G. Andinobates opisthomelas from Guatapé, Antioquia, Colombia, unknown temperature.

BROWN ET AL.

Table 6A (continued on next page). Life History Traits. **= we recognise some of the observations may reflect actual deposition strategies, and could in some cases be the result of cannibalism. Black horizontal line depicts different genera.

Specific epithetet abditus

Mating system -

altobueyensis

Parental care -

Trophic egg feeding -

Clutch number/ deposition sites terrestrial bromeliads (possibly)

Deposition strategy+ individually (possibly)

individually**

bombetes

absent

individually

claudiae

absent

3-4 eggs deposited in leaf litter

1-2 per phytotelm

daleswansoni

male parental male parental -

dorisswansonae

communally, “1-6 larvae in different states… were found inside a bromeliad” -

fulguritus

absent

1-5 eggs

individually

absent

1-4 eggs

opisthomelas

male parental male parental -

tolimensis

viridis

virolinensis

promiscuous

absent

individually

amazonica

promiscuous

male parental male parental

1-6 eggs, at water surface (or just below) in phytotelmata

individually

benedicta

promiscuous

absent or temporally dependent; see account absent

individually

cyanovittata

4-6 eggs, in leaf litter on forest floor -

minutus

male parental -

defleri

individually

fantastica

promiscuous

absent

2-6 eggs

individually

flavovittata

male parental -

individually

imitator

monogamous

1-3 eggs deposited vertically above phytotelm 1-4 eggs; away from water, > 14 cm above pool. 1-4 eggs in leaf litter, often hidden. 1-3 eggs per clutch, deposited on inner wall, above the water, of bamboo internodes or other phytotelmata. 4-9 eggs; in leaf litter on forest floor 1-2 eggs, above phytotelm, ca. 5-20 cm. above pool 2-5 eggs at surface of (or just above) water level in phytotelm or away from phytotelm in leaf litter, often hidden. 1-2 eggs, above phytotelm

reticulata

biparental care male parental male parental

present

male parental -

absent

sirensis

promiscuous

summersi

promiscuous

toraro

uakarii

promiscuous

male parental

absent

vanzolinii

monogamous

present

variabilis

promiscuous

biparental care male parental

ventrimaculata

yavaricola

male parental -

TAXONOMIC REVISION OF RANITOMEYA

absent

individually individually; occasionally, multiple in single phytotelm individually

individually individually Individually; occasionally, entire clutch is abandoned in single phytotelm individually

2-6 eggs at surface of water level (or just below) in phytotelm -

individually, if not, entire clutch is abandoned in single phytotelm. -

individually

Table 6B (continued from previous page). Life History Traits.

Specific epithet

Additional notes

Citation

abditus

Myers & Daly 1976

altobueyensis

Species ID uncertain, possibly minutus

Silverstone 1975

bombetes

Myers & Daly 1980

claudiae

High territoriality, presumed to be strictly terrestrial, observed transporting 1-3 tadpoles observed transporting 1-2 tadpoles

daleswansoni

dorisswansonae

Observed transporting 1-2 tadpoles

Jungfer et al. 2000; unpub. data Thomas Ostrowski Rueda-Almonacid et al. 2006 -

fulguritus

Silverstone 1975

minutus

Silverstone 1975, unpub. data Thomas Ostrowski

opisthomelas

tolimensis

Bernal-Bautista et al. 2007

viridis

virolinensis

Predominantly terrestrial, observed transporting 1-2 tadpoles, no territoriality observed, single bromeliads could contain up to 5 individuals. In French Guiana populations, researchers observed this species change its food-provisioning strategy; early in breeding season, tadpoles subsist predominantly on food within phytotelm, whereas later in the breeding season (likely due to increased desiccation risk), as a food source parents deliberately deposit embryos in pools containing their offspring to ensure quick metamorphosis.

Ruiz-Carranza & RamírezPinilla, 1992

amazonica

benedicta

This paper, Poelman & Dicke 2007

Brown et al. 2008c

cyanovittata

defleri

Twomey & Brown 2009

flavovittata

Likely territorial, males frequently participate in vocal bouts

this paper

imitator

Extremely territorial, males defend territories using vocal bouts and wrestling.

Brown et al. 2008ab, Brown 2009 this paper

Males territorial, males participate in vocal bouts and wrestling

Von May et al. 2008b; J.L. Brown, unpub. data Brown et al. 2008c

this paper Caldwell & de Oliveira 1999 this paper

ventrimaculata

Extremely territorial, males defend territories using vocal bouts and wrestling. Males transported between 1-6 tadpoles. Little to no site fidelity or territoriality. This species exhibits larval parasitism, where males deposit tadpoles in pools containing non-related embryos that are cannibalized when they hatch and fall into the phytotelm. Males have been observed carrying 1-3 tadpoles at a time.

yavaricola

Likely territorial, males frequently participate in vocal bouts

Perez-Peña et al. 2010

fantastica

Brown et al. 2008c

reticulata sirensis summersi toraro uakarii vanzolinii variabilis

E. H. Poelman, unpub. data

BROWN ET AL.

Definition and diagnosis. SVL 16–19 mm; black to bronze dorsum; conspicuous dorsolateral stripes absent (weak dorsolateral stripes present in A. bombetes); head brightly spotted or uniformly yellow or red with identically colored labial stripe (except in A. abditus); white, yellow, orange or red spots in groin (not upper surface of thigh as in Ranitomeya) and between axilla and upper surfaces of forearms (sometimes extending to elbow); limbs uniformly dark (occasionally dorsal coloration extending to limbs, i.e., some populations of A. opisthomelas); venter black, either uniform (A. abditus), with pale to bright spotting (A. dorisswansonae, A. opisthomelas and A. tolimensis), or marbled bluish-white (A. bombetes, A. virolinensis); LTRF either 2(2)/3 (known in A. bombetes and A. virolinensis) or 2(2)/3(1) (known in A. abditus, A. opisthomelas and A. tolimensis); larvae with medial gap in papillae on posterior labium (known only in A. abditus, A. bombetes, A. opisthomelas, A. virolinensis and A. tolimensis, Table 4); oral disc emarginated. Vocalizations of species in this group characterized as drawn-out buzz or rattle, consisting of notes 0.8–2.3 sec in length, repeated at approximately 6–17 notes per minute (Fig. 8, Table 5)

Ranitomeya Bauer 1988 Account authors: J.L. Brown, E. Twomey Figs. 3, 4, 9–45, Tables 2–7 Type species. Dendrobates reticulatus Boulenger, 1884 “1883” Proposed sister group. Andinobates gen. nov. Definition and diagnosis. Unambiguous synapomorphies include: 5 nuclear and 16 mitochondrial synapomorphies (based upon the dataset used in this study, Fig. 3, Tables 2 and 3); distinctive pale reticulation on limbs and venter present (Fig. 2c-i and Fig. 2h-i). Secondary losses of this pale limb reticulation appear to have occurred in R. yavaricola and R. summersi, some morphs of R. fantastica, R. imitator and R. sirensis. Other features include: adult SVL less than 21 mm; adults typically brightly colored, often with bright yellow, red, or green dorsal coloration (either uniform, spotted, or striped); dorsolateral stripe, if present, extending to top of thigh (vs. not reaching thigh in Andinobates), ventrolateral stripe and oblique lateral stripe present or absent; Distinctive, bright coloration on throat present (usually yellow, orange or red); dorsal skin texture nearly smooth to weakly granular; head narrower than body; teeth absent; vocal slits present in males; finger I greatly reduced and shorter than finger II; finger discs II–IV greatly expanded; in adults disc on finger III at least two times wider than distal end of adjacent phalanx; thenar tubercle conspicuous (commonly vestigial, occasionally absent); toe discs III–V moderately expanded; toe webbing absent; median lingual process absent; larval vent tube dextral; larval oral disc emarginated; larvae without medial gap in papillae on posterior labium (known in all species expect R. cyanovittata and R. ventrimaculata, Table 4); scansorial; adults use arboreal phytotelmata for reproduction and deposit eggs away from or at edge of water in phytotelmata (Table 6); tadpoles deposited individually, typically by male; small clutches (2–6 eggs, Table 6); vertebrae 2 and 3 unfused (known in R. amazonica, R. toraro sp. nov., R. imitator, R. variabilis, R. vanzolinii, R. fantastica, R. reticulata and R. sirensis). Distribution. This genus occurs within Amazonian rainforests of Brazil (States: Acre, Amapá, Amazonas, Pará, Rondônia), Bolivia (Department: Pando), Colombia (Departments: Amazonas, Caquetá, Putumayo (tentative), Vaupés), Ecuador (Provinces: Morona-Santiago, Napo, Orellana, Pastaza, Sucumbíos), French Guiana (Arrondissements: Cayenne, Saint-Laurent-du-Maroni), Guyana (District: Potaro-Siparuni) and Peru (Departments1: Amazonas, Cusco, Huánuco, Junín, Loreto, Madre de Dios, Pasco, San Martín, Ucayali). Species within this genus occur between sea level and 1600 m, Fig. 9. Species included (16). Dendrobates amazonicus Schulte 1999; Ranitomeya benedicta Brown, Twomey, Pepper & Sanchez-Rodriguez 2008; Ranitomeya cyanovittata Perez-Peña, Chavez, Twomey & Brown 2010; Ranitomeya defleri Twomey & Brown 2009; Dendrobates fantasticus Boulenger 1884 “1883”; Dendrobates flavovittatus Schulte 1999; Dendrobates imitator Schulte 1986 with its junior synonyms Dendrobates imitator intermedius Schulte 1999 and Dendrobates imitator yurimaguensis Schulte 1999 (Vences & Lötters 2000, Lötters et al. 2003 1.

Since 2009, Peru officially reclassified ‘Departmentos’ as ‘Regiones’. At the time of writing, however, ‘region’ is infrequently used in scientific literature. Here we use Departments as a synonym of Regions.

TAXONOMIC REVISION OF RANITOMEYA

and this paper); Dendrobates reticulatus Boulenger 1884 “1883” with its junior synonym Dendrobates tinctorius igneus Melin 1941; Dendrobates sirensis Aichinger 1991 with its junior synonyms Dendrobates biolat Morales 1992 and Dendrobates lamasi Morales 1992 (this paper); R. summersi Brown, Twomey, Pepper & Sanchez-Rodriguez 2008; Ranitomeya toraro sp. nov. (this paper); Dendrobates uakarii Brown, Schulte & Summers 2006; Dendrobates vanzolinii Myers 1982; Dendrobates variabilis Zimmermann & Zimmermann 1988; Dendrobates ventrimaculatus Shreve 1935 with its junior synonym Dendrobates duellmani Schulte 1999 (this paper); Ranitomeya yavaricola Perez-Peña, Chavez, Twomey & Brown 2010. Remarks. Our definition of Ranitomeya is essentially equal to the definition of Caldwell & Myers (1990) ventrimaculatus group. The genus apparently diverged from Andinobates approximately 14 mya during the mid-Miocene (Santos et al. 2009).

FIGURE 9. Known elevation distributions of Ranitomeya. Dotted line is mean for all samples. Dark boxes display the total elevation range of each species, within each contains a corresponding box plot.

Species groups contained within Ranitomeya Ranitomeya defleri species group Figs. 3, 4, 9, 10–14, Tables 4–7 A monophyletic assemblage of two divergent species: Ranitomeya defleri Twomey & Brown 2009 and R. toraro sp. nov. (this paper). Definition and diagnosis. Adult SVL 13–18 mm; black dorsum; largely complete yellow dorsolateral and middorsal stripes present; cream spots near body on upper surfaces of legs and forearms; limbs and venter black with pale to bright blue reticulation forming round black spots on limbs and irregular spots on venter; large intestine entirely pigmented; LTRF either 2(2)/2(1) (R. toraro sp. nov.) or 2(2)/3(1) (R. defleri), oral disc emarginated (Table 2); larvae gray; eggs dark (Table 6). Vocalizations known only in R. defleri and consist of a series of short, insect-like buzzes, notes 0.41–0.62 sec in length, repeated at approximately 27 notes per minute (Table 5).

BROWN ET AL.

FIGURE 10. Ranitomeya Plate 1. defleri group: A–B: Ranitomeya defleri (all from Vaupés, Colombia); A: Holotype at MCZ (Ω); B: near Estación Biológica Caparú (1 Φ). C–M: Ranitomeya toraro from Brazil; C–D: Careiro da Varzea, Amazonas (A. P. Lima); E: Humaitá, Amazonas (P. I. Simoes); F–G: Cachoiera do Jirau, Rondônia (W. Hödl); H: near Boca do Acre, Amazonas (PRMS and MBS, Ω); I: Host plant of R. toraro: Phenakospermum guyanense near Boca do Acre, Acre (MBS); J: near Boca do Acre, Amazonas (PRMS and MBS); K: Habitat of R. toraro near Boca do Acre, Acre, inset: Aechmea sp. used for tadpole deposition (MBS); L: near Boca do Acre, Amazonas (PRMS and MBS); M: Rio Ituxi, Amazonas (JPC, Ω). (nΦ = number of individual in phylogeny, Ω = population sampled in phylogeny).

TAXONOMIC REVISION OF RANITOMEYA

FIGURE 11. Ranitomeya Plate 2. defleri group: A–H: Ranitomeya toraro (all from Brazil); A-B: Rio Branco, Acre (T. Grant); C-F: Upper Jurua, Acre (unknown). From Colombia: G: Leticia, Amazonas (Jose Manuel Padial, Ω); H: Axil of Aechmea sp. with two R. toraro embryos, near Boca do Acre, Amazonas, BZ (MBS). I: Adelphobates quinquevittatus, near Boca do Acre, Amazonas, BZ (PRMS); J: R. uakarii near Porto Walter, Acre, BZ (JPC, Ω); K: Tadpole of R. toraro (MBS). (Ω = population sampled in phylogeny).

Ranitomeya toraro Brown, Caldwell, Twomey, Melo-Sampaio & Souza sp. nov. Account authors: J.L. Brown, J.P. Caldwell, E. Twomey, P.R. Melo-Sampaio, M.B. Souza Figures 3, 4, 9, 10 (c-m), 11 (a-g, k), 12, 13, Table 4–7 Dendrobates quinquevittatus (non Steindachner 1864) – Silverstone 1975 (partim): p. 35 [CAS 85681 collected by F. Baker and W.M. Mannat “Manaos, Brazil” 1911] Dendrobates ventrimaculatus sensu lato (non Shreve 1935) – Caldwell & Myers 1990 (partim): p. 18, Fig. 11A [MZUSP 63792 and USNM 266119 collected by A. L. Gardner at Cachoeira Nazaré, Brazil, 9 43’S, 61 55”W, on the Rio Jiparaná (= Rio Machado), 1986 (MZUSP specimen not seen by authors)]; Christmann 2004 (partim): p.21, Figs. on p. 21, 99, 153; Brown et al. 2006 (partim): p. 49, Table 2, Figs. 1, 2e, 4; Noonan & Wray 2006 (partim): p. 1012, Table 2, Figs. 2, 4, 5. Ranitomeya ventrimaculata sensu lato (non Shreve 1935) – Grant et al. 2006 (partim): p. 171; Twomey & Brown 2008 (partim): p. 129, Fig. 6; 2009: p. 58; Souza 2009 (partim): p. 32; Perez-Peña et al. 2010 (partim): p. 18, Fig. 13. Ranitomeya sp. – Lötters et al. 2007: p. 510, Fig. 639; Twomey & Brown 2009: p. 58 as sp. B

Holotype. MPEG 13838, an adult male (Fig. 12) collected by J.P. Caldwell (original field number JPC 15713) on 22 December 1998, in Brazil, Amazonas state, municipality of Castanho, at km 12 on road to Autazes (ca. 40 km south of Manaus), 40 m elevation, 3° 30' 52.24" S, 59° 49' 51.13" W. This individual was hopping along a 2 cm horizontal branch about 25 cm above ground in primary forest. Paratypes. All from the state of Amazonas, Brazil (Fig. 12). OMNH 37438–37442, MPEG 13839–13842, collected from 20 Dec 1998 to 11 January 1999 by J. P. Caldwell, L. J. Vitt, S. S. Sartorius, A. P. Lima, M. C. Araújo and T. C. Avila-Pires, same locality as holotype. OHMH 36666–36667 and MPEG 13036–13037, collected from 10 February–26 March 1997 by J. Caldwell, L. J. Vitt, S. S. Sartorius, T. C. Avila-Pires and M. C. Araújo, at Scheffer Madeireira on Rio Ituxi, ca. 170 km southwest of Lábrea, 107 m elevation, 8° 28' 45" S, 65° 42' 59" W.

BROWN ET AL.

FIGURE 12. Ranitomeya toraro sp. nov. type series. All specimens from two localities in Brazil: Amazonas state, municipality of Castanho, at km 12 on road to Autazes (ca. 40 km south of Manaus) or Scheffer Madeireira on Rio Ituxi, ca. 170 km southwest of Lábrea (labeled with ‡). Top row, from the collections of MPEG (L-R): 13839, 13838 (holotype), 13841, 13840, 13842 and 13037(‡). Bottom row, from the collections of OMNH (L-R): 37441, 37440, 36666 (‡), 37442, 37439, 36667(‡), 37438. Black bar = 20 mm (5 mm increments). Sequenced individuals (number in phylogeny): OMNH 36666 (7), OMNH 37440 (5), OMNH 36667 (6), MPEG 13841 (4)

Etymology. The species name is in reference to the noun “to ‘raro,” which in the Apurinã language means “frog.” This indigenous Amazonian tribe occupies the center of the range of the new species. The specific epithet is used as a noun in apposition. Definition and diagnosis. Assigned to the genus Ranitomeya due to the combination of the following characteristics: Adult SVL 15–17.0 mm, dorsal coloration conspicuous, dorsolateral stripes extend to top of thigh, ventrolateral stripes present, brightly colored throat, distinctive pale reticulation on limbs and venter, dorsal skin smooth, finger I greatly reduced and shorter than finger II, finger discs II–IV greatly expanded, disc of finger 1.7–2.3 times wider than finger width, thenar tubercle conspicuous, toe discs III–V moderately expanded, toe webbing absent, larval vent tube dextral, adults use arboreal phytotelmata for reproduction and deposit eggs above phytotelm, maxillary and premaxillary teeth absent. Dorsal body and head black with complete yellow middorsal and dorsolateral stripes, extending from vent to rostrum and extending from upper surface of thighs to orbits, respectively. Limbs and venter black with pale to bright blue reticulation forming round black spots on limbs and irregular spots on venter. Chin yellow with large black paired gular spots and a central spot, irregular in some individuals. Tadpole gray, ovoid, with irregular yellow markings present from early in development. Ranitomeya toraro can be distinguished from all other dendrobatids by its distinctive color pattern. Other species with which it could be confused include R. amazonica, R. defleri, R. flavovittata, R. sirensis (and its junior synonyms R. biolat and R. lamasi), R. uakarii and R. ventrimaculata (including its junior synonym R. duellmani). Ranitomeya defleri and R. flavovittata have highly variable dorsal markings consisting of yellow dots and elongated spots on a black ground color (versus complete middorsal and dorsolateral stripes in R. toraro); further, R. defleri has characteristic large yellow blotches behind the eyes. Ranitomeya flavovittata typically has a pale conspicuous yellow spot on the upper surface of each thigh (versus absent in R. toraro). Some populations of Ranitomeya sirensis have complete, broad, yellow middorsal and dorsolateral stripes. Because of these characters, these morphs are similar in appearance to Ranitomeya toraro; however, in R. toraro the stripes are thinner and the middorsal stripe widens anteriorly, becoming at least two times wider on the head (versus equal width in R. sirensis). Almost all individuals of R. sirensis possess a large spot that is the same color as

TAXONOMIC REVISION OF RANITOMEYA

the stripe coloration (typically yellow) in the center of the venter (versus spot absent in R. toraro). Ranitomeya variabilis and R. amazonica are fairly variable species, but the morphs most similar to R. toraro have a yellow dorsum with a conspicuous black ‘Y’ that extends from the rump to the eyelids (versus the appearance of two black stripes due to complete medial stripe that extends to vent in R. toraro). In southwest Brazil, the co-occurring R. uakarii appears to be a Müllerian mimic of R. toraro, but instead of having bicolored dorsolateral and ventrolateral stripes, all its stripes are completely yellow (see R. uakarii account for complete discussion). Ranitomeya toraro can be distinguished from the southern morphs of R. uakarii by the following characters: absence of a large black spot on the rostrum (versus typically present in R. uakarii), absence of a complete labial stripe (versus present in R. uakarii), fusing of black nostril spots, creating the appearance of a upside-down ‘U’ on the tip of the snout (versus absent in R. uakarii), single pair of large black gular spots (versus typically two smaller pairs in R. uakarii), and a middorsal stripe that is two times wider on the head compared to the sacral region (versus constant width in R. uakarii). Measurements (in mm) of holotype. SVL 15.7; FL 6.9; TL 6.7; KK 12.6; FoL 6.7; HaL 4.4; HL 5.0; HW 5.0; BW 5.5; UEW 2.4; IOD 2.0; IND 1.9; TD 0.9; ED 2.0; DET 0.4; L1F 1.6; L2F 2.4; W3D 0.9; W3F 0.4. For paratype measurements see Table 7. Description of holotype. Head widest at jaw articulation; head slightly narrower than body. Head width equal to head length; head width 31.8% of SVL. Snout acuminate in lateral view, truncate with slight rounding in dorsal view. Naris directed posterolaterally, 1.6 from tip of snout; internarial distance 2.4, 47.2% of head width. Canthus rostralis rounded, loreal region flat. Eye–naris distance 1.4, 70.0% of horizontal eye diameter. Tympanum slightly oval, posterodorsal margin hidden by depressor muscle, tympanum 45.0% of eye diameter. Tongue ovoid, attached anteriorly; median lingual process absent. Teeth absent. Paired vocal slits small, located near jaw articulation. Body with three parallel stripes; middorsal stripe bifurcates anterior to eyes, dorsolateral stripe extends above the center of the eye to groin; ventrolateral stripe integrates into ventral reticulate pattern. Skin texture smooth on head and most of dorsum, becoming weakly granular on posterior surface of dorsum and on limbs. Ventral surfaces of body and limbs weakly granular, chin nearly smooth. Forelimbs slender, hand relatively large, 27.8% of SVL. Finger I considerably shorter than finger II when fingers appressed; finger III > II > IV > I. Discs on fingers II, III and IV greatly expanded; disc on finger I small, rounded. No webbing or lateral fringes on hand. Width of disc on finger III 2.3 times width of adjacent phalanx. Outer metacarpal tubercle (= palmar tubercle) large, round, unpigmented; inner metacarpal tubercle (= thenar tubercle) oval, located at base of finger I. Unpigmented proximal subarticular tubercles present at bases of fingers II, III and IV; on finger I tubercle is halfway to tip of digit. Smaller, distal subarticular tubercles present only on fingers III and IV, tubercle diffuse on finger IV. Tubercles raised in lateral view. Dorsal scutes present on all digits. Length of legs moderate; heels of appessed legs reach level of eyes. Femur and tibia nearly equal in length; tibia 97.3% of femur; knee–knee distance 80.1% of SVL. Relative lengths of appressed toes, IV > III > V > II > I. Toe I short with rounded disc; toes II, IV and V with moderately expanded discs. Unpigmented outer metatarsal tubercle round, raised; unpigmented inner metatarsal tubercle oval, located at base of toe I. Unpigmented outer metatarsal tubercle located laterally at base of fifth metatarsal; unpigmented inner metatarsal tubercle located medially near base of toe I. Weakly defined tarsal keel extends from inner metatarsal tubercle for about one-third length of tarsus; tarsal tubercle absent. Two subarticular tubercles present on toes III and V, one on toes I and II. Three subarticular tubercles on Toe IV, the basal tubercle weakly defined. Toes lack webbing and lateral fringes. Variation. Based on the 14 specimens in the type series, SVL (in mm) of adult males averaged 15.1 ± 0.2 (14.8–15.7; n = 9), adult females 16.4 ± 0.3 (16.2–16.7; n = 2), females thus slightly larger (1.3) larger than males. SVL (in mm) of juvenile females averaged 12.2 ± 0.9 (10.3–13.1; n = 3). The striping pattern among these 14 specimens was remarkably consistent with only minor deviations. Five of the 14 individuals had a slight break in the middorsal stripe; one of these specimens had three breaks and a horizontal stripe connecting the middorsal and left dorsolateral stripe. Only one specimen had a break in either dorsolateral stripe. Specimens from the Autazes locality (n = 10) differed slightly from the Ituxi specimens (n = 4). Those from the Ituxi locality had breaks in the bifurcation of the middorsal stripe, two with breaks on both sides, and two others with breaks only on the left side. No Autazes specimens had breaks in the bifurcation of the middorsal stripe. The dorsolateral stripes in the Autazes specimens were complete (extended below eye to side of body) in two specimens, including the holotype, broken below the eye on both sides in five specimens, and on one side only in three specimens. All Ituxi specimens had breaks below the eye on both sides. Also see variation in Figs. 10 (c-m) and 11 (a-g, k).

BROWN ET AL.

TABLE 7. Measurements of the type series of Ranitomeya toraro. Individual ID: 1= OMNH 3666; 2= OMNH 36667; 3= OMNH 37438; 4=OMNH 37439; 5= OMNH 37440; 6 =OMNH 37441; 7=OMNH 37442; 8=MPEG 13036; 9=MPEG 13037; 10= MPEG 13838; 11=MPEG 13839; 12=MPEG 13840; 13=MPEG 13841, 14=MPEG 13842. All measurements are in mm. ID

SVL

16.2

15.1

13.1

14.9

14.8

16.7

15.6

10.3

15.2

15.47

14.9

13.1

14.8

15.3

Mean (max- min) 14.9 (16.7–13.1)

7.2

7.0

6.0

6.8

8.1

7.0

7.3

4.7

7.3

6.9

5.9

5.5

7.1

6.8 (8.1–5.5)

6.8

6.4

6.0

6.6

7.2

7.0

4.5

6.9

6.7

5.9

5.6

6.9

6.5

6.6 (7.2–5.6)

13.6

13.0

11.0

13.0

14.2

13.6

13.3

---

13.3

12.6

12.1

10.9

14.0

12.6

12.9 (14.2–10.9)

FoL

6.5

6.1

5.4

6.2

6.1

6.6

6.3

3.9

6.4

6.7

5.1

5.7

6.0

6.1 (6.7–5.1)

HaL

4.2

3.7

4.2

4.5

4.3

2.7

4.3

4.4

4.2

3.7

4.1

4.2

4.2 (4.6–3.7)

5.5

4.7

4.5

4.4

5.0

5.5

5.4

3.9

4.9

5.0

4.7

4.2

4.8

4.7

4.9 (5.5–4.2)

5.3

5.1

4.2

4.7

5.0

5.1

3.8

4.9

5.0

4.7

4.4

4.9

4.9 (5.3–4.2)

5.5

5.4

4.6

5.0

5.3

6.0

5.3

4.0

5.2

5.5

5.4

5.0

5.3 (6.0–4.6)

UEW

2.4

2.6

2.3

2.1

2.2

2.7

2.3

1.7

2.3

2.4

2.0

2.4

2.3

2.4 (2.7–2.0)

IOD

2.1

2.0

2.2

2.0

2.1

2.3

1.8

2.4

2.0

2.1

1.9

2.3

2.0

2.1 (2.4–1.9)

IND

2.2

2.1

1.9

2.2

2.1

1.8

2.3

1.9

2.0

2.3

2.1

2.1 (2.3–1.9)

0.9

0.7

0.8

0.9

0.8

0.6

0.7

0.9

0.8

0.9

1.0

0.8 (1.0–0.7)

2.3

2.2

2.0

1.6

2.0

2.6

2.1

1.5

2.2

2.0

1.9

1.7

2.0

1.9

2.0 (2.6–1.6)

DET

0.4

0.6

0.3

0.4

0.3

0.4

0.3

0.4

0.4 (0.6–0.3)

L1F

1.7

1.6

1.3

1.7

1.4

1.3

1.2

1.4

1.6

1.4

1.3

1.6

1.4

1.5 (1.7–1.3)

L2F

2.4

2.3

2.0

2.4

2.3

2.4

2.2

1.9

2.1

2.4

2.0

2.3

2.3 (2.4–2.0)

W3D

0.9

0.8

0.7

0.9

0.7

0.4

0.8

0.9

0.8

0.7

0.8 (0.9–0.7)

W3F

0.4

0.3

0.4

0.3

0.4

0.3

0.4

0.4 (0.4–0.3)

SEX

9M, 2F, 3JF

Color in life. Ituxi specimens: ground color of dorsum black with three longitudinal stripes; dorsolateral stripes are iridescent yellow; middorsal stripe light yellow or light blue on back, becoming deep yellow on head. Middorsal stripe bifurcates on head (incomplete in some specimens) and extends anterior to eye not fusing with labial stripe, dorsolateral stripe extends above the center of the eye (in some speciments posterior to eye) to groin, ventrolateral stipe near axilia is light blue or greenish blue, then yellow in groin. Stripe expands on proximal upper arm to form a triangular light yellow spot. Small yellow rostral spot on tip of snout. Arms, legs and ventral surfaces with blue reticulate pattern on black ground color, forming large black round or elongate spots. Iris black. Specimens from Autazes similar, although some specimens have a light yellow reticulate pattern on the legs and venter. Color in preservative. Dorsum dark gray with three whitish but distinct longitudinal stripes on body and head. Middorsal stripe bifurcates on head anterior to eye and extends below eye alongside to groin. Large central whitish spot on tip of snout. Reticulate pattern distinct, in shades of gray, on upper limbs, sides and all ventral surfaces except chin. Chin with central irregular whitish spot that connects to reticulate pattern on chest. Tadpole. Five back-riding tadpoles were available for description. Three were taken from OMNH 36667 on 13 February and two were taken from MPEG 13037 on 26 March. Both nurse frogs were males and taken from the same locality; therefore, the five tadpoles are described together. All measurements (mm) are means ± SE (range in parentheses). All tadpoles in stage 25; no external gills, not yet feeding, intestine with some yolk remaining. Snout rounded in dorsal and lateral profile; body depressed. Total length 9.5 ± 0.07 (9.3–9.6); body length 3.4 ± 0.04 (3.2–3.5); tail length 6.1 ± 0.05, 64.2% of total length. Body width 2.5 ± 0.1 (2.1–2.6); body depth 1.8 ± 0.09 (1.6–2.1); body depth 72% of body width. Eye well developed; naris small; distance from naris to anterior edge of eye 0.4 ± 0.02 (0.3–0.5). Eye positioned dorsally on head, directed dorsolaterally. Spiracle sinistral; vent tube dextral. Tip of tail bluntly rounded. Tail muscle height at base of tail 0.9 ± 0.04 (0.8–1.0); tail muscle width at base of tail 0.9 ± 0.03 (0.8–1.0); maximum tail height 1.4 ± 0.03 (1.3–1.5). Dorsal fin slightly higher than ventral fin.

TAXONOMIC REVISION OF RANITOMEYA

Oral disc ventral, emarginate; transverse width 0.9 ± 0.02 (0.8–0.9), 36% of body width. Single row of small papillae present laterally and ventrally; dorsal gap where papillae absent. The three tadpoles from OMNH 36667 have only A-1 developed on upper labium; P-1 on lower labium with narrow medial gap; P-2 short. The two tadpoles from MPEG 13037 have A-2 developed in addition to A-1. A-2 composed of two small segments with a large medial gap. LTRF for latter two tadpoles 2(2)/2(1), additional posterior row may develop in older tadpoles.

FIGURE 13. Known distribution of the defleri group. The inset map displays the geographic extent of distributions (black circles = all other Ranitomeya).

Color in preservative. Dorsum light yellow-brown with minute brown flecks. Skin opaque, gut reddishbrown. Tail musculature yellow-brown with tiny brown evenly spaced flecks; tail fins transparent with tiny brown evenly spaced flecks. Vocalizations. Unknown. Distribution and natural history. Ranitomeya toraro occurs in southwestern Brazil and the southeastern tip of Colombia, where it is known from 12 localities (Fig. 13; genetic data are from three localities). This species has been observed in the Brazilian states of Amazonas, Acre and Rondônia and the Colombian department of Amazonas. It likely occurs widely throughout the Madeira, upper Juruá and upper Purus river drainages, although further sampling is needed to determine the extent of its distribution. Individuals from all localities have been found in undisturbed “terra firme” primary forest or old-growth secondary forest in Amazonia. These forests are not subject to flooding during the rainy season. At three localities (Boca do Acre, Autazes, Ituxi), undisturbed forests were characterized by large rainforest trees, such as Brazil nut trees (Bertholletia excelsa). Canopy height varied from 20 to 35 meters. The understory was open to relatively dense. These frogs were relatively uncommon at all sites surveyed. At the Rio Ituxi site, only five frogs were observed in two months of fieldwork. Eleven individuals were found in 24 days at the Autazes site, and nine individuals were found in six days at the Boca do Acre site. At the Rio Ituxi site, frogs were found only in terra firme forest, even though extensive work was done in flooded forests along the river. Of the five frogs observed, two were in leaf litter and three were an average of 35 cm above ground on a leaf, a fallen branch and a log. At the Autazes site, two frogs were found in undisturbed primary forest and nine were found in old second-growth forest. Three were in leaf litter and eight were above ground (average, 80 cm) on leaves (two), logs (three), Heliconia leaves (two) and a dead palm frond (one). To date, tadpole development sites have been found only at the Boca do Acre site. Forests at the Autazes and Ituxi sites had relatively few obvious types of phytotelmata that could serve as tadpole habitats, although Heliconia

BROWN ET AL.

stands were present at the Autazes site. Four frogs found transporting tadpoles on their backs had one, two, three and three tadpoles, suggesting that males frequently carry more than one tadpole at a time. Two egg clutches were found at the Boca do Acre site, one with two eggs 5 cm about the waterline of a bromeliad, Aechmea cf. bromeliifolia and another with a single egg in the axil of the bananeira-brava plant (Strelitziaceae: Phenakospermum guyanense). Five tadpoles were found in bromeliad tanks and 20 were found in axils of P. guyanense; these tadpoles were always found occurring in separate tanks or axils, indicating that cannibalism may occur. Notes. Unlike many species of Ranitomeya, R. toraro displays remarkably little morphological variation across its expansive geographic range. Although this species is sister to R. defleri, the two species display distinct morphological differences and deep genetic divergence. These differences are maintained even when examining R. toraro from Leticia, CO, which is the site where the two species come in closest contact (to our knowledge). Thus, it is unlikely that the morphological and genetic differences we observe between the two species are products of isolation-by-distance. Conservation status. Following the IUCN Red List categories and criteria (IUCN 2010), we tentatively suggest listing this species as Least Concern (LC). The total geographic extent of this species exceeds 800,000 km2 and much of the forests in the Madeira, upper Juruá and upper Purus drainages remain intact. However, given dramatic deforestation rates in Brazil, significant habitat loss in the near future could endanger this species.

Ranitomeya defleri Twomey & Brown 2009 Account author: J.L. Brown Figs. 3, 4, 9–11, 13, 14 Tables 1, 4–6 Dendrobates quinquevittatus (non Steindachner 1864) – Silverstone 1975 (partim): p. 20, Fig. 14, pattern F Ranitomeya defleri Twomey & Brown 2009: p. 1, Figs. 1–6 [MCZ 28061 (holotype) collected by Isadore Cabrera at “Rio Apaporis, Colombia”, 1952] Dendrobates defleri – Santos et al. 2009, by implication

FIGURE 14. Advertisement calls of Ranitomeya species in the variabilis group and defleri group. A. Ranitomeya amazonica from 23 km S Iquitos, Loreto, Peru (type locality), recorded at 26° C; B. Ranitomeya amazonica from French Guiana, unknown temperature (call courtesy Erik Poelman); C. Ranitomeya variabilis from Cainarachi valley, San Martín, Peru, recorded at 22° C. D. Ranitomeya variabilis from Cerro Yupatí, Amazonas, Colombia, recorded at 27° C; E. Ranitomeya variabilis from Saposoa, San Martín, Peru, recorded at 24.5 C; F. Ranitomeya defleri from Rio Apaporis, Vaupés, Colombia, recorded at 26° C.

TAXONOMIC REVISION OF RANITOMEYA

Background information. No new information is available on this species. For a summary of current knowledge on this species see Twomey & Brown (2009). Distribution. This species is known to occur in two localities in the Amazonian rainforests of Vaupés and Amazonas departments, Colombia, though it likely also occurs in Amazonas, Brazil (Fig. 13).

Ranitomeya reticulata species group Figs. 3, 4, 9, 15–18 Tables 1, 4–6 A monophyletic assemblage of six species: Ranitomeya reticulata Boulenger 1884 “1883” with its junior synonym R. ignea Melin 1941; R. fantastica Boulenger, 1884 “1883”; R. ventrimaculata Shreve 1935; R. uakarii Brown, Schulte & Summers 2006; R. summersi Brown, Twomey, Pepper & Sanchez-Rodriguez 2008 and R. benedicta Brown, Twomey, Pepper, & Sanchez-Rodriguez 2008. Definition and diagnosis. Medium to large adult SVL (17–21 mm); black dorsum; typically orange to red dorsal patterning; LTRF 2(2)/3, oral disc emarginate; large intestine entirely pigmented; eggs dark; promiscuous mating system, male parental care; grayish tadpoles and embryos, females produce between 2–5 eggs per mating. Vocalizations consist of a series of very short buzz-like notes (0.1–0.5 sec in length) given in rapid succession (104–200 notes per minute) (Fig. 18).

Ranitomeya benedicta Brown, Twomey, Pepper and Sanchez-Rodriguez 2008 Account authors: J.L. Brown, E. Twomey, M. Pepper Figs. 3, 4, 9, 15 (a–d), 18, 19 Tables 1, 4–6 Dendrobates quinquevittatus (non Steindachner 1864) – Silverstone 1975 (partim): p. 33, Fig. 14 (drawing), patterns K, L [USNM 127933, 127197, 127198 from “Orellana, Domo Santa Clara, Loreto, Peru”, 1947] Dendrobates fantasticus (non Boulenger, 1884 “1883”) – Schulte 1999 (partim): p.57, Fig. 5, patterns K, L; Lima et al. 2006: p. 21, Fig. B Ranitomeya benedicta Brown, Twomey, Pepper and Sanchez-Rodriguez 2008: p. 3, Figs. 1, 3, 5, 11 [MUSM 26957 (holotype) collected by Mark Pepper and Evan Twomey from near Shucushuyacu, Loreto, Peru, 2006]; – von May et al. 2008a: p. 395, Appendix 2 Dendrobates benedictus – Santos et al. 2009, by implication

Background information. For a summary of current knowledge on this species see Brown et al. 2008c. Distribution. This species occurs within the Amazonian rainforests of San Martín and Loreto, Peru (Fig. 19).

BROWN ET AL.

FIGURE 15. Ranitomeya Plate 3. reticulata group: A–D: Ranitomeya benedicta (all from Peru); A–B: Shucushuyacu, Loreto (1Φ); C-D: Pampa Hermosa, Loreto. E–L: Ranitomeya fantastica (all from Peru); E: Yurimaguas, Loreto; F: near Yumbatos, San Martin; G: Pongo de Cainarachi, San Martin (Ω); H: Cainarachi Valley, San Martin (Ω); I: San Antonio, San Martin (KS); J: Tarapoto, San Martin (Ω); K: Santa María de Nieva, Loreto (K.-H. Jungfer, 1Φ); L: Lower Huallaga Canyon, San Martin (Ω). M & N: Ranitomeya summersi (all from San Martin, Peru); M: Chazuta (3Φ); N: Sauce (Ω). O–R: Ranitomeya reticulata (all from Loreto, Peru); O-P: Iquitos (Ω); Q: Puerto Almendras (PPP); R: Upper Rio Itaya (PPP). (nΦ= number of individual in phylogeny, Ω = population sampled in phylogeny). TAXONOMIC REVISION OF RANITOMEYA

FIGURE 16. Ranitomeya Plate 4. reticulata group: A–G: Ranitomeya reticulata (all from Loreto, Peru); A: Lower Rio Itaya (Ω); B–D: Upper Itaya Drainage; E-F: Rio Sucusari (KS); G: 30 km west of Pevas (MSR). H–M: Ranitomeya ventrimaculata: H: Kapawi, Pastaza, EC (L. Coloma, 7Φ); I: Upper Curaray Drainage, Loreto, PE (PPP); J: Yasuní, Orellana, EC (A. Blasco Z., Ω); K: central Rio Nanay (M. Callegari, 6Φ); L: Holotype at MCZ, Sarayacu, Pastaza, Ecuador ; M: Yasuní, Orellana, EC (J. Yeager, Ω); N: Yasuní, Orellana, EC (S. Ron, Ω). O & P: Ranitomeya uakarii (all from Loreto, Peru): O: Tamshiyacu village (8Φ) P: Quebrada Blanco (12Φ). (nΦ = number of individual in phylogeny;Ω = population sampled in phylogeny).

BROWN ET AL.

FIGURE 17. Ranitomeya Plate 5. reticulata group: A–I: Ranitomeya uakarii: A: Central Rio Yavari, Loreto, Peru (PPP); B: Quebrada Blanco, Loreto, Peru (PPP, Ω); C: Rio Boncuya, Loreto, Peru (G. Gagliardi); D: Tournavista, Huánuco (A. Toebe); E: Porto Walter, Acre, Brazil (JPC); F: Rio Los Amigos, Madre de Dios, Peru (RVM); G: R. uakarii sp. aff. Caquetá, Colombia (J. M. Rengifo); H: R. uakarii sp. aff. Iwokrama, Guyana (unknown photographer); I: Guzmania bromeliad with R. uakarii embryos near Tamshiyacu village, Loreto, Peru. (Ω = population sampled in phylogeny).

Ranitomeya fantastica (Boulenger 1884 “1883”) Account authors: J.L. Brown, E. Twomey Figs. 3, 4, 9, 15 (e–l), 18, 20 Tables 1–6 Dendrobates fantasticus Boulenger 1884 “1883”: p. 635, Plate 57, drawing 3 [NHML 1947.2.15.1–4 (four syntypes) collected by Paul Hahnel from “Yurimaguas, Huallaga River, Peru”]; – Myers 1982: p. 1; Kneller 1983: p. 148; Divossen 1999: p. 58, 2002: p. 20; Schulte 1999 (partim): p. 57, Fig. 5, pattern I, L, M Cordillera Oriental “Alto Cainarachi”, Cordillera Oriental “Achinamisa”, Huallaga River “Reticulated Hybrid?” morphs, (I, L, M, reprinted from Silverstone, 1975); Symula et al. 2001 (partim): p. 2415, Fig. 1 photo E, Fig. 3 (phylogenetic tree/drawing); Symula et al. 2003 (partim): p. 452, Table 1, Fig. 3 (phylogenetic tree/drawing); Christmann 2004: p. 6, Figs. on p. 37, 159; Santos et al. 2009, by implication Dendrobates quinquevittatus (non Steindachner 1864) – Silverstone 1975 (partim): p. 33, Fig. 14 (drawing), patterns I, L, M Ranitomeya fantastica – Bauer 1988: p. 1; Grant et al. 2006: p. 171, Fig. 76; Lötters et al. 2007 (partim): p. 472, Figs. 588, 590; Brown et al. 2008a: p. 1154; 2008b: p. 5, Fig. 1; 2008c: p. 2, Figs. 1, 3, 5, 8–10; von May et al. 2008a: p. 394, Appendix 1

Background information. For a summary of current knowledge on this species see Brown et al. (2008c), with the exception of the following comments. Recently, E. Twomey rediscovered a population of R. fantastica that exactly matched Boulenger’s description (identical to 3 of 4 types, Fig 15e) less than 20 km from Boulenger’s stated type locality. Two other populations of R. fantastica are known to occasionally lack black crown markings (near Varadero, Loreto and near Yumbatos, San Martin), however both populations bear subtle differences and occur further from the presumed type locality (ca. 40 km NW and 55 km SW, respectively).

TAXONOMIC REVISION OF RANITOMEYA

In 2008, Karl-Heinz Jungfer discovered a new population of this species in the Cordillera Campanquis near the Pongo de Manseriche, Loreto, Peru. Given its unique morphology and apparently disjunct distribution, we originally suspected that these frogs might represent an undescribed species. However, after performing phylogenetic analyses on sequences collected from these individuals, there is little phylogenetic support for this hypothesis and they appear to simply represent a northern population of R. fantastica. Ironically, the discovery of this population (which is morphologically similar in appearance to R. summersi), further supports the classification of R. summersi as a separate species from R. fantastica. Based on our phylogeny, a population of R. fantastica that occurs less than 30 km from a population of R. summersi is most closely related to a population of R. fantastica from Pongo de Manseriche. These two populations of R. fantastica are separated by more than 250 km and several mountain ranges, versus less than 30 km and relatively contiguous habitat. This relationship could also be attributed to incomplete lineage sorting and should be further investigated. The phylogenetic topology of R. fantastica, R. summersi and R. benedicta differs from that published in Brown et al. (2008c). In this study, the lower Huallaga populations of R. fantastica are sister to R. summersi, which form a clade that is sister to R. benedicta; this entire clade is sister to the rest of the members of R. fantastica. In Brown et al. (2008c), R. summersi was sister to R. fantastica (in which, R. fantastica consisted of two main clades, one containing individuals from the lower Huallaga and an individual from Tarapoto and all individuals from lower Huallaga sister to all other R. fantastica) and both species were sister to R. benedicta. To clarify these differences, we performed reciprocal topology tests (using Shimodaira-Hasegawa tests, Table 3); each study’s topology was tested on both datasets (note: samples not included in both analyses were left as they were in the unconstrained topology as much as possible). When using the current dataset, which contains fewer base pairs than Brown et al. (2008c), we found no statistical differences in either topologies (p = 0.491) suggesting they are equally probable. The differences in tree length (under Parsimony) were 1. However when using the dataset from Brown et al. (2008c), the Shimodaira–Hasegawa test rejected the topology observed in this study (p = 0.012), supporting the specific relationships observed in Brown et al. (2008c), with individuals from lower Huallaga being members of R. fantastica). The difference in tree lengths (via Parsimony) between these two tree topologies was 13. Thus, despite the topology of our phylogeny, we maintain that R. benedicta is sister to a clade containing the sister species R. fantastica and R. summersi. Even though these results support the specific status of the lower Huallaga individuals as R. fantastica (Brown et al. 2008c; Fig. 15), this population’s phylogenetic status (and specific status) is not entirely clear. In other unpublished analyses (based on different datasets), these populations were observed sister to R. summersi (as in this study), and based on morphology we cannot reject this possible relationship (that these populations are members of R. summersi). To clarify this, additional morphological and sequence data are necessary, using both mitochondrial and nuclear genes, and molecular data should be analyzed using coalescent-based phylogenies. Distribution. This species is known to occur within the departments of Amazonas, Loreto and San Martín, Peru (Fig. 20).

Ranitomeya reticulata Boulenger, 1884 “1883” Account authors: J.L. Brown, E. Twomey, S. Lötters, P. Perez-Peña Figs. 3, 4, 9, 15 (o–r), 16 (a–g), 18, 21 Tables 1, 4–6 Dendrobates reticulatus Boulenger, 1884 “1883”: p. 635, Plate 57, drawing 2 [NHML 1947.2.15.5–12 (eight syntypes) collected by Paul Hahnel from “Yurimaguas, Huallaga River, Peru”]; – Myers & Daly 1980: p. 20; Zimmermann & Zimmermann 1984: p. p. 35, 1985; Almendariz 1987: p. 77; Hermann 1988: p. 78; Wiltenmuth & Nishikawa 1994: p. 57; Divossen 1999: p. 58, 2000: p. 20; Rodriguez & Duellman 1994: p. 16; Lötters et al. 2003: p. 1909; Christmann 2004: p. 6, Figs. on p. 87, 92, 96; Santos et al. 2009, by implication Dendrobates tinctorius igneus Melin, 1941: p. 66, Fig. 37A–B [MHNG 19.1.1925, 20.1.1925 (two syntypes) from "Rio Itaya (near Iquitos), Perú", collected in 1925] Dendrobates quinquevittatus (non Steindachner 1864) – Silverstone 1975 (partim): p. 33, Fig. 14 A–C; Lescure & Bechter 1982: p. 26 Ranitomeya reticulata—Bauer 1988: p. 1; Grant et al. 2006: p. 171; Lötters et al. 2007: p. 489, Figs. 619, 620; von May et al. 2008a: p. 394, Appendix 1; Werner et al. 2010, 2011: p. 16, Figs. 1–3 Ranitomeya ignea—Grant et al. 2006: p. 171 Dendrobates igneus—Santos et al. 2009, by implication

BROWN ET AL.

Background information. Ranitomeya reticulata shares, with R. fantastica, the distinction of being the first species of Ranitomeya to be described (Boulenger 1884 “1883”). This species is also the type species of the genus Ranitomeya (Bauer 1988; for discussion see Grant et al. 2006). In the last 30 years, little controversy has surrounded the validity of this species. For a comparison of R. reticulata to R. ventrimaculata, see the R. ventrimaculata account. Tadpole. The description is based on a single live tadpole from Yanamono, Loreto, Peru. Mouthparts were verified in another tadpole from the same locality. The tadpole was feeding on detritus. Tadpole in stage 30; no external gills. Body ovoid in dorsal view, wider near vent. Total length 16.6; body length 9.8; tail length 6.8, tail length 41% of total length. Body width 5.1; body depth 4.0, 78% of body width. Eye well-developed; naris small; distance from naris to anterior edge of eye 0.7. Eye positioned dorsally on head, directed dorsolaterally. Spiracle well developed; vent tube dextral. Tip of tail bluntly rounded. Tail muscle height at base of tail 2.5; tail muscle width at base of tail 2.3; maximum tail height 4.1. Dorsal fin same height as ventral fin. Oral disc ventral, emarginate; transverse width 2.4, 14% of body width. Single row of small papillae present laterally and ventrally; wide dorsal gap where papillae absent. LTRF 2(2)/3(1) with A-1 developed on upper labium, A-2 with wide medial gap (30% of total width); P-1 on lower labium with narrow medial gap; P-2 equal in width to P-1; P-3 80% width of P-1. In life, head and body gray; mouthparts visible from above. Abdomen mostly transparent; intestinal coils black, heart visible. Tail musculature uniform gray, dorsal and ventral fins opaque gray. Natural history. Some authors have observed this species engaging in biparental care and maternal egg provisioning in captivity (Zimmermann & Zimmermann 1984; Christmann 2004), such as has been observed in the wild in R. imitator and R. vanzolinii (Caldwell 1997, Caldwell & de Oliveira 1999, Brown et al. 2010). These behaviors have never been confirmed in the field in R. reticulata, despite considerable attention by researchers (Divossen 1999, 2000; Werner et al. 2011). In these studies, the authors observed a species with male-only parental care and a polygamous mating system. Lastly, other members of the reticulata group (where parental care behaviour is known) demonstrate male-only parental care and biparental care is limited to some members of the vanzolinii group (Summers & McKeon 2004; Lötters et al. 2007). The phylogenetic placement of this species, sister to other male parental care species, cannot exclude the possibility of the independent evolution of this trait; however, it is not likely, given that biparental care/maternal provisioning is only known to have evolved twice in the family Dendrobatidae (Summers & McKeon 2004). Taxonomic notes. In 1941, Melin described Dendrobates tinctorius igneus. Little information is published on this putative taxon, possibly because of a lack of credence in its validity (see Silverstone 1975). Schulte (1999) regarded it as a junior synonym of R. reticulata, and stated that it was a morph that maintained juvenile coloration. Grant et al. (2006) elevated this subspecies to specific status as Ranitomeya ignea, without comment or justification. After examining the holotype, Lötters & Vences (2000) suggested this species may be conspecific with either amazonica and/or reticulata (both sensu this paper). However, based on our (J.L. Brown, E. Twomey, unpub. data) observations on Puente Itaya frogs (the type localiy of ignea), we are confident the frogs described by Melin are referable to reticulata and not amazonica. The holotype (NHMG 512) possesses dorsolateral and middorsal stripes (partially broken) that extend about three-fourths of the SVL (14.9 mm), a characteristic common in striped R. reticulata. In contrast, in R. amazonica the middorsal stripe typically terminates between the shoulders (i.e., about one-fourth of the SVL). Lastly, we observed several populations of R. reticulata along the Iquitos–Nauta road, especially near Puente Itaya, that possess black dorsal markings similar to the holotype of ignea (which forms a dashed ‘U’). As a conclusion, we place Dendrobates tinctorius igneus Melin, 1941 as a junior synonym of Ranitomeya reticulata (Boulenger, 1884 “1883”). Distribution. This species in known to occur within the Amazonian rainforests of Peru (Loreto Department) and Ecuador (Pastaza Province), Fig. 21. Conservation status. Following the IUCN Red List categories and criteria (IUCN 2010), we tentatively suggest listing this species as Near Threatened (NT). Although the distribution of this species is estimated to be around 20,000 km2, much deforestation has occurred along Río Amazonas. Further, Iquitos, a very large city, occupies the center of its distribution.

TAXONOMIC REVISION OF RANITOMEYA

FIGURE 18. Advertisement calls of Ranitomeya species in the fantastica group. A. Ranitomeya reticulata from Iquitos, Loreto, Peru, recorded at 29° C; B. Ranitomeya ventrimaculata, unknown locality or temperature; C. Ranitomeya fantastica from Cainarachi valley, San Martín, Peru, recorded at 24° C; D. Ranitomeya summersi from Sauce, San Martín, Peru, recorded in captivity at 24.5° C; E. Ranitomeya benedicta from Shucushuyacu, Loreto, Peru, recorded in captivity at 26.5° C; F. Ranitomeya uakarii from Rio Tahuayo, Loreto, Peru, recorded in captivity at 26° C.

Ranitomeya summersi Brown, Twomey, Pepper & Sanchez-Rodriguez 2008 Account authors: J.L. Brown, E. Twomey, M. Pepper Figs. 3, 4, 9, 15 (m, n), 18, 19 Tables 1, 4–6 Dendrobates quinquevittatus (non Steindachner, 1864)—Silverstone 1975 (partim): p. 33, Fig. 14 (drawing), patterns G, H Dendrobates fantasticus (non Boulenger, 1884 “1883”)—Schulte 1999 (partim): p. 57, Fig. 5 pattern G, H, Cordillera Oriental “West flank” and Cordillera Azul “North spur”; Symula et al. 2001 (partim): p. 2415, Fig. 1 photo E, Fig. 3 (phylogenetic tree/drawing); 2003 (partim): p. 458, Table 1 (D. fantasticus, Sauce), Fig. 3 (phylogenetic tree/drawing; D. fantasticus, Sauce); Christmann 2004: p. 32, Figs. on p. 32 Ranitomeya fantastica (non Boulenger, 1884 “1883”)—Grant et al. 2006 (partim): p. 171; Lötters et al. 2007 (partim): p. 473, Fig. 592 Ranitomeya summersi Brown, Twomey, Pepper & Sanchez-Rodriguez, 2008: p. 9, Figs. 1, 5, 6, 10. [MUSM 26994 (holotype) collected by Jason L. Brown and Evan Twomey near the town of Sauce, San Martín, Perú];—von May et al. 2008a: p. 396, Appendix 2 Dendrobates summersi—Santos et al. 2009, by implication

Background information. For a summary of knowledge on this species see Brown et al. (2008c). An individual from near Chazuta (the locality of some of the paratypes) was included in this study’s phylogeny. The results support the classification of this population based on morphology as R. summersi (Brown et al. 2008c). See R. fantastica account for discussion of this study’s phylogenetic results and the results of Brown et al. (2008c). Distribution. This species in known to occur within a small area of San Martín, Peru (Fig. 19).

BROWN ET AL.

FIGURE 19. Known distribution of Ranitomeya benedicta and R. summersi. The inset map displays the geographic extent of distributions (black circles = all other Ranitomeya).

FIGURE 20. Known distribution of Ranitomeya fantastica. The inset map displays the geographic extent of distributions (black circles = all other Ranitomeya).

TAXONOMIC REVISION OF RANITOMEYA

Zootaxa 3083 漏 2011 Magnolia Press 路

FIGURE 21. Known distribution of Ranitomeya reticulata and R. ventrimaculata. The inset map displays the geographic extent of distributions (black circles = all other Ranitomeya).

FIGURE 22. Known distribution of Ranitomeya uakarii. The inset map displays the geographic extent of distributions (black circles = all other Ranitomeya).

56 路 Zootaxa 3083 漏 2011 Magnolia Press

BROWN ET AL.

Ranitomeya uakarii Brown, Schulte & Summers 2006 Account authors: J.L. Brown, E. Twomey, R. von May, J.P. Caldwell, P.R. Melo-Sampaio, M.B. Souza Figs. 3, 4, 9, 16 (o, p), 17(a–h), 18, 22 Tables 1, 4–6 Dendrobates quinquevittatus (non Steindachner 1864)—Silverstone 1975 (partim): p. 33 Dendrobates ventrimaculatus (non Shreve 1935)—Rodriguez & Knell 2003 (partim): p. 148; Christmann 2004 (partim): p. 19, Figs. on p. 19, 97, 154; Brown et al. 2006 (partim): p. 45, Table 2, Figs. 1, 4; Roberts et al. 2006a (partim): p. 382, Table 1, Fig. 4 Dendrobates duellmani (non Schulte 1999)—Roberts et al. 2006a (partim): p. 377, Table 1, Figs. 1, 4. Dendrobates uakarii Brown, Schulte & Summers 2006: p.47, Figs. 1–4, Table 1, 2 [MUSM 23246 (holotype) collected by Mario Callegari upstream Quebrada Blanco near the Tamshiyacu–Tahuayo Reserve, Department Loreto, Peru, 2004];— Santos et al. 2009, by implication Ranitomeya uakarii—Grant et al. 2006, by implication; Lötters et al. 2007: p. 494, Fig. 624; von May et al. 2008a: p. 396, Appendix 2; Perez-Peña et al. 2010: p. 2, Figs. 8, 13 Ranitomeya cf. reticulata (non Boulenger, 1884 “1883”)—Lötters et al. 2007: p. 493, Fig. 623. Ranitomeya cf. uakarii—Lötters et al. 2007: p. 495, Fig. 625 Ranitomeya ventrimaculata (non Shreve 1935)—Perez-Peña et al. 2010 (partim): p. 18, Fig. 13

Background information. Prior to the formal description of this species as R. uakarii, it was considered a morph of R. ventrimaculata sensu this paper (see below). The type series consists of five individuals collected from just outside Tamshiyacu–Tahuayo reserve. Since this description, we have learned of additional populations of uakariilike frogs from numerous localities throughout the Amazon basin. Molecular phylogenetic analyses indicate that these populations represent several morphs not discussed in the original description, making it necessary to redefine R. uakarii. Definition and diagnosis. Assigned to the genus Ranitomeya due to the combination of the following characters: Size small (adults < 17.5 mm SVL), dorsal coloration conspicuous, dorsolateral stripes extend to top of thigh, ventrolateral stripes present, brightly colored throat, distinctive pale reticulation on limbs and venter, dorsal skin smooth, finger I greatly reduced and shorter than finger II, finger discs II–IV greatly expanded, disc of finger 2–2.5 times wider than finger width, thenar tubercle conspicuous, toe disc III–V moderately expanded, toe webbing absent, larval vent tube dextral, adults use arboreal phytotelmata for reproduction and deposit eggs above the waterline in phytotelm, maxillary and premaxillary teeth absent. Tadpole gray, ovoid, with irregular markings present from early in development. Three morphs of this species are currently known: (i) the Nominotypical morph (Figs. 16o–p, 17a–c, h), (ii) the Toraro morph (Fig. 17e, f) and (iii) the Tri-Country morph (Fig. 17g). The limbs and venters of all morphs are reticulated in blue to greenish blue on black, forming black spots. (i) The Nominotypical morph has a dorsum with paired red or reddish-orange dorsolateral stripes (thin or thick) that extend from the tip of the snout, where they form a U-shape, to the groin. A single middorsal stripe starts on the top of the head and terminates above the vent. This stripe may occasionally form a ‘Y’ on the head, connecting with the dorsolateral stripes to create a black spot on top of snout. A yellow labial stripe continues to form a spot on the upper forearm. A yellow, irregular ventrolateral stripe is present. Venter is reticulated blue or greenish blue on black. Chin is yellow with paired black spots. This morph occurs throughout the lowlands east of Río Ucayali and southeast of Río Amazonas in Peru (Loreto and possibly Ucayali Departments) and has been observed in central Peru along Río Pachitea (Huánuco Department). A population of frogs similar in appearance has been discovered in Guyana (District: Potaro-Siparuni, Fig. 17h). Due to the seemingly widespread distribution of R. uakarii, we provisionally consider the Guyana population as R. uakarii. (ii) The Toraro morph is so named because of its strong resemblance to R. toraro sp. nov. (in fact, these two species were long thought to be conspecific and may in fact be Müllerian mimics, see mimicry section). It has thin yellow dorsolateral and oblique lateral stripes. A similar middorsal stripe begins at the top of the head and terminates above the vent. The stripe typically divides into a ‘Y’ on the head (between the orbits) connecting with both dorsolateral stripes. The ventrolateral stripes often form irregular connections to the dorsolateral stripes. This morph is infrequently encountered, though it can be locally abundant (e.g., near Porto Walter, Brazil). This morph occurs within the lowland forests of southwestern Brazil (States: Acre, Amazonas), in southeastern Peru (Department: Madre de Dios) and possibly into northwestern Bolivia (Department: Pando). TAXONOMIC REVISION OF RANITOMEYA

(iii) In the Tri-Country morph, both the dorsolateral and middorsal stripes blend with the limb and flank reticulation in the middle of the dorsum (between the axilla and groin, gradually changing from red to blue). The black space between the dorsolateral and middorsal stripe typically creates the appearance of a black ‘U’ with the bottom near the snout (occasionally this ‘U’ is broken and the red nose spot fuses with the red middorsal stripe). The upper lip has a short yellow stripe that terminates on the upper surface of the forearms. This morph lacks a distinct oblique lateral stripe. This morph has not been genetically sampled, thus its classification should be considered provisional. This rare morph is known to occur in southeastern Colombia (Departments: Amazonas and Caquetá) near the three-country corner shared between Peru, Colombia and Brazil. Thus, we would not be surprised if this morph were discovered in nearby Brazil or Peru. The Nominotypical morph is similar in appearance to Ranitomeya amazonica, R. ventrimaculata (sensu this paper) and some morphs of R. reticulata. Ranitomeya amazonica typically possesses a conspicuous black ‘Y’ which starts mid-dorsum and terminates at the snout (some morphs of R. ventrimaculata appear similar to this, possessing a broken ‘U’, giving the appearance of a black ‘Y’, but they typically have complete to largely complete middorsal stripes). Individuals of R. uakarii with red dorsolateral stripes typically have yellow ventrolateral stripes (versus typically absent in R. ventrimaculata and R. reticulata, or, if present, the ventral and limb reticulation are the same color). The Toraro morph can be distinguished from its apparent mimic, R. toraro, by the following characters: presence of a large black spot on the rostrum (typically absent in R. toraro), presence of a complete labial stripe (absent in R. toraro), non-fused black nostril spots (nostril spots fused in R. toraro, creating the appearance of a upside down ‘U’ on the tip of the snout), typically two smaller pairs of black gular spots (single pair of large spots in R. toraro), and a middorsal stripe that is constant width (middorsal stripe two times wider on the head compared to the sacral region in R. toraro). Often the ventrolateral and dorsolateral stripes are connected by small irregular stripes, middorsally. The Tri-Country morph is similar in appearance to R. reticulata and R. ventrimaculata sensu this paper but possesses a yellow labial stripe (versus red in R. reticulata and pinkish-red in R. ventrimaculata). Tadpole. The description is based on a single live tadpole from Río Tahuayo, Loreto, Peru. Mouthparts were verified in two other tadpoles of R. uakarii from Quebrada Blanco, Loreto and Tournavista, Huánuco. The tadpole was feeding on detritus. Tadpole in stage 29, no external gills. Body ovoid in dorsal view, wider near vent. Total length 16.5; body length 6.3; tail length 10.2, 62% of total length. Body width 4.6; body depth 3.7, 80% of body width. Eyes well developed; naris small, distance from naris to anterior edge of eye 0.6. Eye positioned dorsally on head, directed dorsolaterally. Spiracle well developed; vent tube dextral. Tip of tail bluntly rounded. Tail muscle height at base of tail 1.8; tail muscle width at base of tail 1.6; maximum tail height 2.5. Dorsal fin slightly higher than ventral fin. Oral disc ventral, emarginate; transverse width 1.6, 35 % of body width. Single row of small papillae present laterally and ventrally; dorsal gap where papillae absent. LTRF 2(2)/3(1) with A-1 developed on upper labium, A-2 with wide medial gap (one-third total width of tooth row); P-1 on lower labium with narrow medial gap; P-2 equal width of P-1; P-3, 75% width of P-1. Color in life. Head gray, mouthparts visible from above. Body gray, color evenly distributed; abdomen mostly transparent, intestinal coils black, heart visible. Tail musculature uniform gray, dorsal and ventral fins opaque gray. Distribution. This species occurs in Amazonian rainforests of Brazil (States: Acre, Amazonas), possibly into Bolivia (Department: Pando), Colombia (Departments: Amazonas, Caquetá), Guyana (Potaro-Siparuni) and Peru (Departments: Huánuco, Loreto, Madre de Dios and possibly Ucayali), Figure 22. Conservation status. Following the IUCN Red List categories and criteria (IUCN 2010), we suggest listing this species as Least Concern (LC). It exhibits a large geographical range and occurs within several protected areas. Taxonomic remarks. This species warrants further study and could possibly represent a species complex given the geographically distinct morphs. Under the current phylogeny we find little support for this arrangement and the nominotypical and Toraro morph are not reciprocally monophyletic.

BROWN ET AL.

Ranitomeya ventrimaculata Shreve 1935 Account authors: J.L. Brown, E. Twomey, E.H. Poelman Figs. 3, 4, 9, 16 (h–n), 18, 21 Tables 1, 4–6 Dendrobates minutus ventrimaculatus Shreve 1935: p. 213 [MCZ 19734 (holotype) collected by O. C. Felton from “Sarayacu, Ecuador” 1933] Dendrobates quinquevittatus (non Steindachner 1864)—Silverstone 1975 (partim): p. 33; Almendariz 1987 (partim): p. 77; Lötters 1987 (partim) p. 72, Fig. 1 Dendrobates ventrimaculatus—Daly et al. 1987 (partim): p. 1025; Rodriguez & Duellman 1994 (partim): p. 17, Plate 2d; Caldwell & Myers 1990 (partim): p. 1; Santos et al. 2009, by implication Dendrobates duellmani Schulte, 1999: p. 69 [NHMK 231832 (holotype, wrongly given as NHMK 221832 in the original description and corrected by Lötters & Vences 2000: p. 252), unspecified collector from “San Jacinto, 2 km, nahe der ekuadadorianischen Grenze, Loreto, Peru,” restricted to “Südöstliche Umgebung vom Ölcamp San Jacinto (bei 2°18’42.8“ S, 75°51’57.9“ W), circa 180 m NN, östlich des Rio Tigre, Departamento Loreto, Peru” by Lötters & Vences 2000: p. 252];—Lötters & Vences 2000: p. 252; Santos et al. 2003: p. 12794, Fig. 1; Cisneros-Heredia 2003; Christmann 2004: p. 8, Figs. on p. 16, 103, 104; Darst & Cannatella 2005: p. 60, Fig. 1; Brown et al. 2006: p. 45, Table 2, Figs. 1, 2; Roberts et al. 2006a (partim): p. 377, Table 1, Figs. 1, 4; Santos et al. 2009, by implication Dendrobates sp. G—Santos et al. 2003: p. 12794, Fig. 1; Darst & Cannatella 2005: p. 60, Fig. 1 Ranitomeya aff. uakarii—Lötters et al. 2007: p. 496, Fig. 626 Ranitomeya duellmani—Grant et al. 2006: p. 171; Lötters et al. 2007: p. 470, Fig. 587: Venegas & von May 2010: p. 282; von May & Mueses-Cisneros 2011: p. 306.

Background information. This species was originally described by Shreve (1935) as a subspecies of Dendrobates minutus. Silverstone (1975) considered it a junior synonym of Dendrobates quinquevittatus. In 1982, Myers wrote that “New material, as well as closer attention to morphological details and evidence of sympatry, convinces me that Dendrobates quinquevittatus Steindachner, sensu Silverstone is a composite of five or more species of distinctively colored frogs.” Interestingly, almost a decade later Caldwell & Myers (1990) would point out that none of those species were in fact D. quinquevittatus sensu stricto. In 1999, Schulte described Dendrobates duellmani from northern Peru. Its description promptly sparked criticism from Lötters & Vences (2000), primarily because most of the description was based on a photograph of a single live frog (from Rodriguez & Duellman 1994), which is traceable as NMHK 231832, and the use of other unorthodox “specimens” including a postal stamp issued by the Ecuadorian government. Nevertheless, due to unique morphology, Lötters & Vences (2000) agreed that duellmani was a valid species. In 2006, Grant et al. placed this species in the genus Ranitomeya within the ventrimaculata group. That same year, Roberts et al. (2006a) demonstrated that duellmani was polyphyletic, with one individual (the population closest to the type locality) nested within a clade of reticulatus and another individual that was sister to a species referred to as ventrimaculatus (the latter now considered to be Ranitomeya uakarii, see account in this paper). Recently, the plea of Caldwell & Myers (1990) that taxonomists proceed with caution when dealing with the systematics of this group was further justified. After reading Shreve’s description of Dendrobates minutus ventrimaculata and examining the type material, J.L. Brown was surprised to learn that most of the specimens, including the holotype, were actually identical to R. duellmani. It is not clear exactly when the “unofficial” definition of R. ventrimaculata changed to R. variabilis sensu this paper; however, shortly after the description of R. duellmani, this “definition” became absolute in the scientific literature. One cause for this change was likely due to the presumption that Schulte (1999) considered the type specimens of R. ventrimaculata when describing duellmani. The situation was further exacerbated by an abundance of observations of the more widely distributed species: variabilis and amazonica sensu this paper, which at that time were also being called ventrimaculata. Due to the clarity of Shreve’s definition (see excerpt below) and morphology of the type specimens, we herein formally consider Ranitomeya ventrimaculata to be senior synonym of the junior synonym Ranitomeya duellmani. For classification of species considered to be R. ventrimaculata from 1987 to 2010, see the R. variabilis and R. amazonica accounts.

TAXONOMIC REVISION OF RANITOMEYA

Black above, three prominent, longitudinal, grayish lines extending from the head almost the entire length of the back, the outermost starting about the posterior border of the upper eyelid, the median about on a level with the anterior border of the upper eyelid (in the paratypes [and type] these dorsal lines are often rather pinkish). Fifteen of the paratypes have substantially the same dorsal markings as the type; ten [now considered to be R. variabilis] show an alternative set of dorsal markings. In these the median dorsal line is much shortened and sends out a pair of branches on the head (sometimes one of the pair is missing)…[and] the outermost lines [i.e., the oblique and dorsolateral lines] are often joined by a transverse line at their posterior ends [commonly observed in R. variabilis but not R. ventrimaculata sensu stricto]… B. Shreve 1935, p. 214

Definition and diagnosis. Assigned to the genus Ranitomeya due to the combination of the following characters: adults small, SVL < 18.0 mm, dorsal coloration conspicuous, dorsolateral stripes extending to top of thigh, brightly colored throat, distinctive pale reticulation on limbs and venter, dorsal skin smooth, finger I greatly reduced and shorter than finger II, finger discs II–IV greatly expanded, disc of finger two times wider than finger width, thenar tubercle conspicuous, toe discs III–V moderately expanded, toe webbing absent, larval vent tube dextral, maxillary and premaxillary teeth absent. This species possesses thin pink to reddish-pink middorsal and dorsolateral stripes. Dorsolateral stripes typically extend from the groin, pass through the eyes and fuse at the tip of the snout. Middorsal stripe starts between the eyes and typically terminates above the vent. Ventrolateral and oblique lateral stripes absent. In most populations, both the dorsolateral and middorsal stripes blend with the limb and flank reticulation in the middle of the dorsum (between the axilla and groin, gradually changing from red to blue). The black space between the dorsolateral and middorsal stripe typically creates the appearance of a black ‘U’ with the bottom near the snout (occasionally this ‘U’ is broken and the red nose spot fuses with red medial stripe). Labial stripe, identically colored as dorsal stripes, terminates at the upper surface of the forearms. Venter, flanks and limbs are finely reticulated in blue-gray to gray on black; the throat coloration varies from being entirely pinkish to pinkish only at the tip (with the remaining throat being predominantly black). Some populations near Río Nanay, ca. 40 km west of Iquitos, Loreto, and on Río Momón near Iquitos lack a middorsal stripe and possess pinkish to plum dorsolateral stripes. This species is similar in appearance to Ranitomeya amazonica, R. uakarii and some morphs of R. reticulata. Ranitomeya amazonica typically possesses a conspicuous black ‘Y’ that starts mid-dorsum and terminates at the snout (some morphs of R. ventrimaculata appear similar to this, possessing a broken ‘U,’ giving the appearance of black ‘Y;’ however, they typically have a complete to largely complete middorsal stripe). Ranitomeya uakarii individuals with red dorsolateral stripes typically have yellow oblique lateral stripes (typically absent in R. ventrimaculata, or, if present, they are the same coloration as ventral and limb reticulation). Some preserved specimens lose their red dorsal coloration, and can appear identical to some morphs R. uakarii. The latter generally possess distinct gular spots that do not connect, giving the appearance of an hourglass on the throat (versus fused gular spots, possessing light coloration- in life pinkish- only at the anterior edge of the throat in R. ventrimaculata). Certain morphs of R. reticulata (and almost all juveniles of this taxon) possess a similar black ‘U,’ although it is typically broken rather than a complete stripe. The middorsal stripe is broad, blending with incomplete dorsolateral stripes that terminate near the mid-dorsum (Fig. 16a, f). The Río Nanay/Río Momón populations are distinct; no other Ranitomeya lacks a middorsal stripe while possessing pinkish dorsolateral stripes (Fig. 16k). Natural history. Individuals of this species were found in undisturbed primary forest with deep leaf litter and low to moderately dense understory vegetation. In Napo province, Ecuador, we observed 1 to 3 individuals per day, suggesting that these frogs occur in low densities. Most frogs were observed foraging on leaf litter, occasionally climbing on tree trunks up to a meter above the forest floor. At localities where they were observed, bromeliads were not abundant on the forest floor (in many cases absent), but large bromeliads were present in the canopy. On several occasions we found vocalizing males, and males carrying 2 or 3 tadpoles on the forest floor. These observations suggest that the species has male parental care and egg clutches may be deposited in the leaf litter. However, to date, we have not been able to locate egg clutches or where males deposit their tadpoles. Ranitomeya variabilis, R. reticulata, R. uakarii, Allobates zaparo, Ameerega bilinguis and A. parvula co-occur with R. ventrimaculata. Vocalizations. The call of this species consists of a rapid buzz, notes 0.32–0.38 sec in length, repeated at approximately 124 notes per minute. Interestingly, in both R. ventrimaculata and its sister species R. reticulata, pulse rates are remarkably high compared to other members in their species group. Within a note, we

BROWN ET AL.

found that individual ‘pulses’ are actually composed of triplets or quadruplets, a character that was not found in other members of the reticulata species group. This difference may be related to the quality of the recording (being able to detect such fine note structure in high quality recordings) rather than physiological differences in call production between species. Distribution. This species occurs in Amazonian rainforests of Colombia (Amazonas Department), Ecuador (Provinces: Napo, Orellana, Pastaza, Sucumbíos) and Peru (Loreto Department), Fig. 21. Conservation status. Following the IUCN Red List categories and criteria (IUCN 2010), we tentatively suggest listing this species as Least Concern (LC). Taxonomic remarks. Over the past decade, mostly unpublished discussion has questioned the legitimacy of this species. Some have suggested that R. ventrimaculata is a morph of R. reticulata with the retention of juvenile coloration (paedomorphy; Roberts et al. 2006a). We remain skeptical of classifying these species as distinct because the potential distributions of both R. reticulata and R. ventrimaculata overlap considerably, something that is rare among closely related species. Further, in some areas it is impossible to distinguish them from each other, e.g., an individual from Kapawi, Ecuador is genetically classified as R. ventrimaculata, however is identical in coloration and pattern to some populations R. reticulata (possessing broad-bright red dorsal stripes versus thin-pinkish dorsal stripes observed in all other R. ventrimaculata). However, because our phylogenetic results support two reciprocally monophyletic clades (Fig. 3b) that are mostly consistent with the morphological differences ascribed to each species, we suggest they remain distinct until more data are available.

Ranitomeya vanzolinii species group Figs. 3, 4, 9, 23–29 Tables 1, 4–6 A monophyletic assemblage of six species: Ranitomeya vanzolinii Myers 1982; R. sirensis Aichinger 1991 with its junior synonyms R. biolat Morales 1992 and R. lamasi Morales 1992; R. imitator Schulte 1986 with its junior synonyms R. imitator intermedius Schulte 1999 and R. imitator yurimaguensis Schulte 1999; R. flavovittata Schulte 1999; R. yavaricola Perez-Peña, Chavez, Twomey & Brown 2010 and R. cyanovittata Perez-Peña, Chavez, Twomey & Brown 2010. Definition and diagnosis. Medium to large adult SVL (17–21 mm); black dorsum; highly variable morphology; LTRF 2(2)/3(1), oral disc emarginated; large intestine unpigmented; larvae whitish to gray; eggs cream; territoriality present in males; promiscuous (known in R. sirensis) or monogamous mating system (known in R. vanzolinii and R. imitator), male parental care (observed in R. sirensis) or biparental (observed in R. vanzolinii and R. imitator); females produce between 2–4 eggs per mating. The advertisement call of species in this group consists of a loud trill, highly tonal, with notes 0.4–1.1 sec in length, repeated at 2–14 notes per minute (Fig. 28).

Ranitomeya cyanovittata Perez-Peña, Chavez, Brown & Twomey 2010 Account author: J.L. Brown Figs. 3, 4, 9, 23 (a, b), 29 Tables 1, 4–6 Ranitomeya cyanovittata Perez-Peña, Chavez, Brown & Twomey 2010: p. 12, Figs. 8, 10 [CORBIDI 02266 (holotype) collected by Diego Vasquez in the Río Blanco Basin near Zona Reservada Sierra Del Divisor, Departamento Loreto, Peru, September 2008]

Background information. For a summary of current knowledge on this species see Perez-Peña et al. 2010. Our phylogenetic results place this species in the vanzolinii group, sister species to R. yavaricola. Distribution. This species occurs within Amazonian rainforests of Peru (Departments: Loreto, possibly Ucayali) and possibly Brazil (States: Acre, Amazonas), Fig. 29.

TAXONOMIC REVISION OF RANITOMEYA

FIGURE 23. Ranitomeya Plate 6. vanzolinii group: A & B: Ranitomeya cyanovittata: Sierra del Divisor, Ucayali, Peru (G. Knell and D. Vasquez, 1:Ω,2: 1Φ). C & D: Ranitomeya yavaricola (all from Loreto, Peru): C: Rio Blanco (G. Knell); D: Lago Preto (PPP, Ω). E– I: Ranitomeya flavovittata (all from Quebrada Blanco, Loreto, Peru (Photo credits: JLB, ET and PPP, Ω). J–K: Ranitomeya vanzolinii Atalaya, Ucayali, Peru (J. Yeager). L–V: Ranitomeya imitator (All from San Martin, Peru): L–O: Upper Canarachi Valley (‡); P– Q: Tarapoto (‡); R: Shapaja (‡); S: Chumia (‡) and T–V: Chazuta (Ω). (nΦ = number of individual in phylogeny, Ω = population sampled in phylogeny, ‡ = genetically sampled, but not included in our phylogeny).

BROWN ET AL.

FIGURE 24. Ranitomeya Plate 7. vanzolinii group: A–Y: Ranitomeya imitator (all from San Martin, Peru unless noted): A-B: Chazuta (Ω); C: Central Huallaga Canyon (‡); D-H: Callanayacu (‡); I-J: Lower Huallaga Canyon (‡); K-Q: Pongo de Cainarachi (Ω); RS: Balsapuerto, Loreto (‡); T–V: Varadero, Loreto (‡) and W–Y: Curiyacu (‡). (Ω = population sampled in phylogeny, ‡ = genetically sampled, but not included in our phylogeny).

TAXONOMIC REVISION OF RANITOMEYA

FIGURE 25. Ranitomeya Plate 8. vanzolinii group: A–I: Ranitomeya imitator Curiyacu, San Martin, Peru (‡). J–T: Ranitomeya sirensis (all from Peru unless noted): J-L: CICRA Station, Madre de Dios (Rio Los Amigos, Ω); M: near Rio Branco, Acre, Brazil (PRMS); N & O: Central Rio Urubamba, Cusco (G. Chavez); P & Q: Tingo Maria, Huánuco (ET, Ω); R: Bamboo forest, R. sirensis often uses the phytotelmata within bamboo for tadpole deposition, Tingo Maria, Huánuco (ET); S: Aguaytía, Ucayali; T: Codo del Pozuzo, Huánuco (20Φ ). (nΦ = number of individual in phylogeny, Ω = population sampled in phylogeny, ‡ = genetically sampled, but not included in our phylogeny).

BROWN ET AL.

FIGURE 26. Ranitomeya Plate 9. vanzolinii group: A–P: Ranitomeya sirensis (all from Peru): A–G: Puerto Inca, Huánuco (JLB and ET, Ω); H & I: Breeding pair of R. sirensis found in the type locality between the lowland and highland populations, Cordillera El Sira, Huánuco (MSR); J & K: Rio Pachitea, Huánuco (J. Stenicka); L–N: Cordillera El Sira, Huánuco (B. Wilson and JLB, 10-11Φ); O: Rio Pachitea, Huánuco (J. Stenicka, 17Φ); P: Yanayacu Maquia, Ucayali. (nΦ = number of individual in phylogeny Ω = population sampled in phylogeny).

TAXONOMIC REVISION OF RANITOMEYA

FIGURE 27. Ranitomeya Plate 10. vanzolinii group: A–F: Ranitomeya sirensis (all from Peru): A–D: near Contamana, Loreto (JLB and G. Gagliardi, 1-3Φ); E: uncertain locality, likely Iscozacin, Junin (7 Φ); F: Estación Biológica Paujil, Junin (L. Schulte). (nΦ = number of individual in phylogeny).

Ranitomeya flavovittata Schulte 1999 Account authors: E. Twomey, J.L. Brown, P. Perez-Peña Figs. 3, 4, 9, 23 (e–i), 29 Tables 1, 4–6 Dendrobates flavovittatus Schulte 1999: p.80 [CRS BD 10H (holotype) from “Boca des Río Tahuayo, Nordufer, 120 m NN” = mouth of Río Tahuayo, north shore, 120 m.a.s.l, collected by Rainer Schulte];—Lötters & Vences 2000: p. 252; Brown et al. 2006: p. 46, Fig. 1; Roberts et al. 2006a: p. 378, Table 1, Figs. 2, 4; Santos et al. 2009, by implication Ranitomeya flavovittata—Grant et al. 2006: p.171; Lötters et al. 2007: p. 47, Figs. 595, 596; Twomey & Brown 2008: p. 129, Table 3, Fig. 6, 2009: p. 50; von May et al. 2008a: p. 395, Appendix 2; Perez-Peña et al. 2010: p. 2, Figs. 7, 8, 13

Background information. This species was described from a single juvenile reared in the laboratory of INIBICO (Instituto de Investigación Biológica de las Cordilleras Orientales, Tarapoto, Peru). Schulte (1999) suggested that this species was closely related to Adelphobates quinquevittatus and A. castaneoticus, primarily on the basis of larval morphology. Schulte (1999) described the call as follows: “a short, harsh trill-call, this call type is unprecedented in Peru and forms a new call group” [translated from German]. Although the description was lacking in some aspects (particularly with regard to intraspecific variation), Lötters & Vences (2000), in a severe critique of Schulte (1999), nevertheless agreed on the validity of R. flavovittata. Roberts et al. (2006a) published the first genetic data for this species, supporting its validity, and found it to be nested within the vanzolinii group as the sister taxon to R. vanzolinii. Morphologically, R. flavovittata strongly resembles R. yavaricola, although the latter lacks black markings on the legs. Each of the traits appears to be fixed within each species, respectively (Perez-Peña et al. 2010).

BROWN ET AL.

Tadpole. The description is based on a single tadpole in stage 26 from Río Tahuayo, Loreto, Peru. Tadpole in stage 26. Body ovoid in dorsal view, wider near vent. Total length 17 mm; body length 7.3; tail length 9.7, 57% of total length. Body width 6.9; body depth 5.4, 78% of body width. Eye well developed; naris small; distance from naris to anterior edge of eye 1.8. Eye positioned dorsally on head, directed dorsolaterally. Spiracle well developed; vent tube dextral. Tip of tail bluntly rounded. Tail muscle height at base of tail 2.7; tail muscle width at base of tail 1.7; maximum tail height 3.9. Dorsal fin slightly higher than ventral fin. Oral disc ventral, emarginate; transverse width 1.9, 28% of body width. Single row of small papillae present laterally and ventrally; dorsal gap where papillae absent. LTRF 2(2)/3(1) with A-1 developed on upper labium, A-2 with wide medial gap (one-third total width of oral disc); P-1 on lower labium with narrow medial gap; P-3 80% width P-1; P-2 equal to width P-1. Color in life. Head whitish, mouthparts visible from above. Abdomen whitish, mostly transparent, intestinal coils tan, heart visible. Tail musculature uniform tan, dorsal and ventral fins semi-transparent, white. Natural history. Almost nothing has been published on the natural history of this species. Schulte (1999) mentioned that he found two white eggs in a small bromeliad (Guzmania sp.) 1.2 m from the ground. Based on our own observations, this species appears to have similar life-history traits as other members of the vanzolinii group. We have witnessed adults (presumably males) carrying single tadpoles in the vicinity of Guzmania bromeliads. We also encountered one courting pair in the axils of a species of palm. Near Rio Yavari-Mirin, we encountered a single tadpole within a bromeliad attached to a liana 1.5 meters above the ground. Males call regularly throughout the day. It is still unknown whether this species displays biparental care like its closest relatives R. vanzolinii and R. imitator (Caldwell & de Oliveira 1999; Brown et al. 2010). However, anecdotal data from terrarium observations reveal that tadpoles of this species engage in “begging” behavior similar to the behavior displayed by R. imitator and R. vanzolinii when begging for food eggs (Chris Miller, pers. comm.). Therefore, this species may also have biparental care, although further study is required. Vocalizations. The call of R. flavovittata is a stereotypical vanzolinii species group call: loud trill, notes 0.81.1 sec in length, repeated at roughly 2 notes per minute (Fig. 28, Table 5). We disagree with Schulte’s assertion that this species has an unusual call which should be considered a new call group. At the time of this species’ description in 1999, the calls of several species of the vanzolinii species group frogs were known and the call of R. flavovittata fits into that call type. Distribution. The type locality is described as ‘mouth of Río Tahuayo, north shore.’ Although this leaves some ambiguity as to the precise location (Río Tahuayo flows into the Amazon on a north–south axis, therefore the two ‘shores’ are better described as east or west), we can assume that the type locality is south of the Amazon. Since 2004 we have documented several other localities for this species, all of which are in the Tamshiyacu– Tahuayo region, except for two records: one from Río Yavari and another further south from nearby Genaro Herrera (Fig. 29, Giuseppe Gagliardi, pers. comm). Based on extensive field work within this region (mainly by P. Perez-Peña), it appears that R. flavovittata has a highly restricted range, although areas south of Río Yavari remain under-studied. Several flavovittata-like frogs have been observed in Brazil near Reserva Extrativista do Alto Juruá and Parque Nacional da Serra do Divisor; however, it remains undetermined whether these frogs represent populations of R. flavovittata or R. vanzolinii. Surveys of these populations are central to clarifying the alpha-taxonomic status of this species. This species occurs within Amazonian rainforests of Peru (Departments: Loreto, possibly in Ucayali) and possibly within Brazil (States: Acre, Amazonas). Conservation status. The IUCN Red List of Threatened Species considers this species as Data Deficient (DD) on the basis of the unverified taxonomic status and the ambiguous type locality. This species is now known from 6 localities in the Tamshiyacu–Tahuayo region of northern Peru and its taxonomic status appears to be valid. We recommend listing this species as Least Concern (LC), applying IUCN Red List categories and criteria (IUCN 2010) since it is locally abundant and occurs in habitat that is largely unaffected by humans. Taxonomic remarks. As in previous studies (e.g., Roberts et al. 2006a, Twomey & Brown 2008), our results support the arrangement of R. flavovittata as sister species to R. vanzolinii. However, our expanded dataset (which includes five R. flavovittata as opposed to a single individual in previous studies) raises some uncertainty as to the taxonomic status of this species. In our current phylogeny, R. flavovittata is paraphyletic with respect to R. vanzolinii. Still, R. vanzolinii and R. flavovittata are morphologically distinct and therefore we maintain the current tax-

TAXONOMIC REVISION OF RANITOMEYA

onomy here until more data can be collected, particularly from intermediate geographical regions (as mentioned above). Regardless of the aforementioned taxonomic issues, introgressive hybridization appears to have occurred between this species and other members of the vanzolini group; we collected an individual of R. sirensis (Panguana morph from Contamana) whose mtDNA was most closely related to R. flavovittata (suggesting historic hybridization between a female R. flavovittata and a male R. sirensis, Fig. 3b). If R. flavovittata is determined to be a separate species from R. vanzolinii, the observed phylogenetic topology could be attributed to historic introgression between the ancestors of R. flavovittata and R. vanzolinii. However, until additional populations of R. vanzolinii are surveyed, these questions cannot be addressed. For additional information see the R. vanzolinii account.

FIGURE 28. Advertisement calls of Ranitomeya species in the vanzolinii group. A. Ranitomeya vanzolinii from Pongo de Mainique, Cuzco, Peru, recorded in captivity at 26° C; B. Ranitomeya sirensis from Ishanga near Tocache, San Martín, Peru, recorded at 25.5° C; C. Ranitomeya imitator from Varadero, Loreto, Peru, recorded at 23.5° C; D. Ranitomeya imitator from Cainarachi valley, San Martín, Peru, recorded at 22 C; E. Ranitomeya flavovittata from Rio Tahuayo, Loreto, Peru, recorded in captivity at 25° C; F. Ranitomeya yavaricola from Lago Preto, Loreto, Peru, recorded at 24° C (call courtesy Pedro Pérez-Peña).

Ranitomeya imitator Schulte 1986 Account authors: J.L. Brown, E. Twomey Figs. 3, 4, 9, 23 (l–v), 24 (a–y), 25 (a–i), 28, 30 Tables 1, 4–6 Dendrobates quinquevittatus (non Steindachner 1864)—Silverstone 1975 (partim): p. 35; Zimmermann and Zimmermann 1984 (partim): p. 35; 1988 (partim): p. 132 Dendrobates ventrimaculatus (non Shreve 1935)—Caldwell & Myers 1990 (partim): p. 18, Fig. 11 Dendrobates imitator Schulte 1986: p. 11, Figs. 1–10 [MUSM BATR 10501 (holotype) collected by Rainer Schulte at km 33, near the village of San Jose, Carretera Tarapoto–Yurimaguas, San Martín, Peru];—Caldwell & Myers 1990: p. 17, Fig. 11c; Divossen 1999: p. 59; Symula et al. 2001: p. 2145, 2003: p. 452, Table 1, Figs. 1–6; Christmann 2004: p. 6, Figs. on

BROWN ET AL.

p. 26, 27, 123–125; Brown et al. 2008a: p. 1140, Table 1, Fig. 1, 2008b: p. 1, Table 1, Figs. 1–3, 2009b: p. 478, Table 1–4, Fig. 3, 2009c: p. 148; Santos et al. 2009, by implication Dendrobates imitator imitator—Schulte 1999: p. 93 Figs. PB-015, PB-016; Lötters & Vences 2000: p. 253; Lötters et al. 2003: p. 1905; Christmann 2004: p. 142, Figs. on p. 142, 143 Dendrobates imitator intermedius Schulte 1999: p. 95, Figs. ZB-009, PB-037, PB-034, PB-036, PB-029, PB- 040, PB-031 [CRS BD 27 (holotype) collected by Rainer Schulte at “Huallaga Canyon, Department San Martín, Perú, 200 m NN.”];— Lötters & Vences 2000: p. 252; Christmann 2004: p. 30, Figs. on p. 30, 127, 128 Dendrobates imitator yurimaguensis Schulte 1999: p. 104, Figs. DB-056, PB-051, PB-035, PB-039 [CRS BD 41 (holotype) collected by Rainer Schulte on “Carretera Yurimaguas–Tarapoto, Alto Amazonas, Río Paranapura Drainage”];—Christmann 2004: p. 28, Figs. on p. 28, 29. Ranitomeya imitator—Bauer 1988: p.1; Grant et al. 2006: p. 171; Lötters et al. 2007: p. 478, Figs. 597–609; Brown et al. 2008c: p. 9; 2009a: p. 1877, Table 1, 2010: p. 436, Figs. 1, 2, 4, 5; von May et al. 2008a: p. 396, Appendix 2 Ranitomeya intermedia—Grant et al. 2006: p. 171; von May et al. 2008a: p. 396, Appendix 2 Dendrobates intermedius—Santos et al. 2009, by implication

Background information. Few poison frogs have received as much attention from the public and the scientific community as R. imitator. Initial studies were focused on this species’ validity and its mimetic relationship with R. variabilis and R. summersi, each sensu this paper, e.g., Schulte (1986), Symula et al. (2001, 2003). Recently, this species has been the focus of several studies regarding the evolution of parental care, mate choice and the evolution of Müllerian mimicry (e.g., Brown et al. 2008ab, 2009b, 2010). This species was discovered in the late 1980s by Rainer Schulte (1986). After naming the species in 1986, Schulte (1999) further subdivided it into three subspecies, R. imitator imitator, R. imitator yurimaguensis and R. imitator intermedius (described below as three color morphs). These names posed various taxonomic problems (see Lötters & Vences 2000) and immediately after their publication, the name imitator yurimaguensis was formally synonymized with the nominotypical form by Lötters & Vences (2000). However, imitator intermedius tentatively remained. Grant et al. (2006) elevated the status of imitator intermedius to full species (as R. intermedia), but provided no justification for this taxonomic arrangement. Ranitomeya imitator has been the subject of several phylogenetic and population genetic studies and has been densely sampled throughout its known range (Symula et al. 2001, 2003; E. Twomey & J.L. Brown, unpub. data). In addition, researchers have collected a considerable amount of acoustic, morphological and behavioral data on most populations of this species (E. Twomey, J.L. Brown & J. Yeager, unpub. data). None of the data from these studies indicate that R. imitator is a species complex or adequately justifies the use of subspecies. In particular, recent population genetic studies have demonstrated that most color morphs are not reciprocally monophyletic, or even genetically distinct. Furthermore, within many populations different color morphs are known to coexist. To maintain the use of subspecies (and the resulting classification of R. intermedia) is misleading, ignoring considerable variation within and between most populations. Because of these reasons, we consider R. intermedia to be a junior synonym of R. imitator. Definition and diagnosis. Assigned to the genus Ranitomeya due to the combination of the following characters: adult SVL < 20.0 mm, dorsal coloration conspicuous, dorsolateral stripes, when present, extend to top of thighs, brightly colored throat, distinctive pale reticulation on limbs and venter (absent in some populations), dorsal skin smooth, finger I greatly reduced and shorter than finger II, finger discs II–IV greatly expanded, disc of finger two times wider than finger width, thenar tubercle conspicuous, toe discs III–V moderately expanded, toe webbing absent, larval vent tube dextral, adults use phytotelmata for reproduction and deposit eggs away from phytotelm, maxillary and premaxillary teeth absent. Vocal slits present in males. Rantiomeya imitator is one of the most polymorphic poison frog species and a Müllerian mimic throughout most of its range, making identification of this taxon difficult. Furthermore, it possesses no single diagnostic morphological character that is consistent for all morphs. Three primary morphs exist: (i) Striped morph (Fig. 24b–g, i– v), (ii) Spotted morph (Fig. 23l–v, 24a, h) and (iii) Banded morph (Figs. 24w–y, 25a–i). Although these three morphs predominate many populations do not exactly fit any of them (e.g., Fig. 24b, c, x, y). To thoroughly describe the morphs of a species that possesses such an immense variation undervalues complicated intrapopulation and interpopulation variation. Thus, the color morphs described below should be interpreted ‘loosely’ and in the context that they do not always represent a population or common ancestry. (i) The Striped morph has a black dorsum with three thin yellow longitudinal stripes extending the length of the body. The presence of two paired spots on the nostrils creates the appearance of a yellow ‘cross’ anterior to the

TAXONOMIC REVISION OF RANITOMEYA

eyes (occasionally spots fuse to make a black “U,” Fig. 24o). Flanks are black and usually have a single yellow ventrolateral stripe that runs from the axilla to the groin. Legs, forelimbs and venter are black with fine light green to blue reticulation. The striped morph occurs throughout the lowlands northeast of the Cordillera Escalera, extending eastward, across the Río Huallaga, into the Pampas del Sacramento. It occurs as far north as Varadero (Department Loreto) but not reaching Río Marañón, and as far south as the northern Cordillera Azul, San Martín, Peru.

FIGURE 29. Known distribution of Ranitomeya flavovittata, R. yavaricola and R. cyanonvittata. The inset map displays the geographic extent of distributions (black circles = all other Ranitomeya).

FIGURE 30. Known distribution of Ranitomeya imitator. The inset map displays the geographic extent of distributions (black circles = all other Ranitomeya).

BROWN ET AL.

FIGURE 31. Known distribution of Ranitomeya sirensis and R. vanzolinii. The inset map displays the geographic extent of distributions (black circles = all other Ranitomeya).

(ii) The Spotted morph has a dorsal ground color of orange, golden yellow, or green. The dorsum has small to medium black spots that occasionally fuse, resembling irregular stripes. This morph typically possesses paired black spots on the nostrils (occasionally these spots are fused and resemble a ‘U’). Legs, forelimbs and venter are the same as the Striped morph, but can also be finely to coarsely reticulate in the colors of teal to gold. This morph is montane and occurs throughout the Cordillera Escalera as far north as Balsapuerto, Loreto, Peru and as far south as Chazuta, San Martín, Peru. (iii) The Banded morph is the most variable; however, the majority of individuals possess a black dorsum with symmetrical orange (occasionally yellow) bands (occasionally possessing a complete or partial vertebral stripe). This morph is found throughout the central Huallaga canyon, San Martín, Peru. Ranitomeya imitator is a Müllerian mimic of R. summersi (banded morph) and R. variabilis (spotted and striped morphs, see discussion). Ranitomeya imitator can be distinguished from these species by its loud, trill-like call (compared to faint buzz-calls in the model species) which is audible from over 5 meters. It can be distinguished from R. variabilis by the presence of black paired nostril nose spots (single spot in R. variabilis). Ranitomeya imitator can usually be distinguished from R. summersi by the presence of a black ovoid head spot (versus pentagonal in R. summersi) and the absence of paired black gular spots (present in R. summersi). This species, particularly the Striped morph, is also similar in morphology to some morphs of R. sirensis, although R. imitator lacks a yellow ventral patch and white axillary and inguinal spots typical of most forms of R. sirensis. Tadpole. The description is based on a stage 26 tadpole from Cainarachi Valley, San José, San Martín, Peru. Body ovoid in dorsal view, wider near vent. Total length 17.5; body length 6.7; tail length 10.8, 62% of total length. Body width 3.7; body depth 2.2, 60% of body width. Eye well developed; naris small; distance from naris to anterior edge of eye 1.6. Eye positioned dorsally on head, directed dorsolaterally. Spiracle well developed; vent tube dextral. Tip of tail bluntly rounded. Tail muscle height at base of tail 1.7; tail muscle width at base of tail 0.9; maximum tail height 1.9. Dorsal fin slightly higher than ventral fin. Oral disc ventral, weakly emarginate; transverse width 1.4, 38% of body width. Single row of small papillae present laterally and ventrally; dorsal gap where papillae absent. LTRF 2(2)/3(1) with A-1 developed on upper labium, A-2 with wide medial gap (one third total width of A-2); P-1 on lower labium with narrow medial gap; P3, 55% width P-1; P-2 equal in width P-1.

TAXONOMIC REVISION OF RANITOMEYA

Color in preservative. Head whitish, mouthparts visible from above. Whitish abdomen, mostly transparent, intestinal coils black. Tail musculature uniform tan, dorsal and ventral fins opaque white. Natural history. This species is a Müllerian mimic of several congeneric species throughout its range (see discussion). Ranitomeya imitator exhibits a seasonal monogamous mating system and biparental care. After one to two embryos hatch, the male carries the tadpoles on its back and deposits each individually in a phytotelm. Both parents then return weekly and the female deposits trophic eggs after being stimulated by the male. The tadpoles are continually fed until metamorphosis. For details and discussion of these behaviors and their evolution, and the natural history of this species see Brown et al. (2008a, b, 2009b, 2010). Vocalizations. The call of R. imitator is stereotypical for members of the vanzolinii group. It is a loud trill, notes 0.44–1.07 sec in length, repeated at 7–11 notes per minute. Despite major morphological differences between populations, acoustic divergence between populations appears to be minimal (E. Twomey, unpub. data). Distribution. This species occurs within Amazonian rainforests of Peru (Departments: Loreto and San Martín), Fig. 30. Conservation status. Following the IUCN Red List categories and criteria (IUCN 2010), we tentatively suggest listing this species as Least Concern (LC). Its distribution is estimated to be around 10,000 km2. Because of the pet market demand and extreme morphological variation over a small geographic area, some color morphs maybe vulnerable.

Ranitomeya sirensis Aichinger 1991 Account authors: J.L. Brown, E. Twomey, M. Pepper, M. Sanchez-Rodriguez, P.R. Melo-Sampaio, M.B. Souza Figs. 3, 4, 9, 25 (j–t), 26 (a–p), 27 (a–f), 28, 31 Tables 1–6 Dendrobates quinquevittatus (non Steindachner 1864)—Silverstone 1975: p. 35 (partim) [Huánuco, Tingo Maria (CAS 85147, 85150; USNM 1666904-06)]; Meede 1980: p. 39 Dendrobates ventrimaculatus (non Shreve 1935)—Caldwell & Myers 1990 (partim): p. 19, Fig. 12 [USNM 268841–268844 collected by Reginald B. Cocroft at Tambopata, Madre de Dios, Peru] Dendrobates sirensis Aichinger 1991: p. 1, Fig. 1–3, Table 1[NHMW 31892 (holotype) collected by Manfred Aichinger in the Serrania de Sira, Río Llullapichis drainage, Huánuco, Peru, 1987];—Schulte 1999: p. 135, Fig. DB-075; Santos et al. 2009, by implication Dendrobates biolat Morales 1992; p. 195, Table 2, 3, Figs. 3–5 [MUSM 7143 (holotype) collected by Victor R. Morales at the Reserva de la Biosfera del Manu, Madre de Dios, Perú, 1987]; – Duellman & Thomas 1996; Schulte 1999: p. 121, Fig. DB-071; De La Riva et al. 2000: p. 3; Doan & Arriaga 2002: p. 108; Lötters et al. 2003: p. 1908; Christmann 2004: p. 6, Figs. on p. 37, 159; Medina-Müller 2006: p. 1; Roberts et al. 2006a: p. 381, Table 1, Figs. 1, 4; Brown et al. 2008a: p. 5, Fig. 1; Santos et al. 2009, by implication Dendrobates lamasi Morales 1992; p. 191, Fig. 1, 2, 5 Table 1, 2 [MUSM 1461 (holotype) collected by Victor R. Morales at Bosque Castilla, northwest of Iscozacín, Huanacabamba, Pasco, Peru, 1986];—Schulte 1999: p. 115, Figs. PB-050, PB-018; Lötters et al. 2003: p. 1909; Symula et al. 2003: p. 453, Table 1, Figs. 2–6; Christmann 2004: p. 6, Figs. on p. 34, 35, 129–132, 148–149, 155– 157; Roberts et al. 2006a: p. 381, Table 1, Fig. 4; Santos et al. 2009, by implication Ranitomeya biolat—Grant et al. 2006: p. 171, Fig. 76; Lötters et al. 2007: p. 466, Figs. 582–586; Maldonado et al. 2007: p. 14, Fig. 1; von May et al. 2008a: p. 395, Appendix 2, 2008b: p. 66, Table 1, Fig. 1; Waldram 2008: p. 232, Fig. 1-2; Melo-Sampaio & Souza 2009: p. 447 Ranitomeya lamasi—Grant et al. 2006: p. 171, Fig. 76; Lötters et al. 2007: p. 484, Figs. 610–618; von May et al. 2008a: p. 396, Appendix 2 Ranitomeya sirensis—Grant et al. 2006: p. 171; Lötters et al. 2007: p. 513, Figs. 614–642; von May et al. 2008a: p. 394, Appendix 1

Background information. This species was described in 1991 (as Dendrobates sirensis) on the basis of four specimens collected in the 1970s and 1980s from the western slope of the northern Cordillera El Sira (also known as the Serranía de Sira). This site was the focus of several scientific investigations during the 1960s and 1970s, including one herpetological survey conducted in 1973 (Duellman and Toft 1979). Surprisingly, this survey did not encounter any Ranitomeya species and it was not until 1976 that this species was first collected, found in elfin forest at 1,560 m elevation by Werner Hanagarth. Later, in the mid 1980s, Manfred Aichinger led a herpetological survey in the Sira which lasted over a year. During this survey, only five individuals were encountered, two of which escaped after capture.

BROWN ET AL.

Recent efforts have been made to sample this interesting species. Not only was this species described from a remote and isolated mountain range, but morphologically it seemed quite distinct from other Ranitomeya species, lacking the limb reticulation typical of this genus. In 2005, M. Pepper and M. Sanchez-Rodriguez led an expedition to the Sira in an attempt to find R. sirensis. They were able to reach one of Aichinger’s camps but were unable to locate any R. sirensis. In 2007, Pepper and Sanchez-Rodriguez returned to the Sira with J.L. Brown and E. Twomey and were lucky enough to find two adult R. sirensis less than 4 kilometers from the type locality at an elevation of 442 m (an elevation much lower than previously known for this species—it remains questionable whether this species occurs at elevations > 1400 m, as stated in the original description). This discovery occurred while returning to the camp after searching higher elevation forests less than 1 km from the type locality. The two frogs perfectly matched the original description (pictured in Fig. 26l, n) and were wedged deep within the axil of a Xanthosoma plant on an abandoned hunting camp. Later that year J.L. Brown sequenced samples collected from toe-clips of these individuals and was surprised to find the sequences nested within a clade of Ranitomeya lamasi, described by Morales (1992), from Panguana. This clade was part of a larger clade composed of other R. lamasi and R. biolat, also described by Morales (1992). Not satisfied with the knowledge gained from this expedition, Pepper and Sanchez-Rodriguez returned again in 2008 to further clarify the status of this species. When previously in the Sira, the researchers ignored the abundant calls of R. lamasi and did not pursue these individuals. However, in light of the close relationship between lamasi and sirensis based on sequence data, the 2008 team pursued all individuals with lamasi-like calls. After several days the team had collected several courting pairs and solitary individuals. Two of the pairs were partaking in “hybridization” events between R. lamasi and R. sirensis. The first observation consisted of a male R. sirensis (SVL 17.3 mm) courting a female R. lamasi (SVL 15.6 mm, Fig. 26h). After being captured, the pair was held overnight in a plastic bottle (prior to being released) and they bred. The eggs were fertilized and began to develop. The second observation was of a male R. lamasi courting an intermediate female (likely a mix of the two nominal species: it had coloration like R. sirensis with SVL 16.1 mm, though more orange with faint broken black stripes and ventral spotting, Fig. 26m). Lastly a lone individual with intermediate morphology (as described above) was observed. These data suggest that these interactions are not atypical at the site where we observed these two nominal species. In fact, only the two individuals sequenced lacked any dark pigmentation (of the five frogs with morphologies similar to R. sirensis). There is the possibility that the two R. sirensis individuals sequenced were also “hybrids” and both were descendants of parents that had undergone maternal introgression (from R. lamasi) and the two represent natural species. However, given the frequency of hybridization, similar call and the shared ventral spot, we consider it highly likely that sirensis and lamasi represent one species. In 2007, J.L. Brown spent four days in southern Peru at the CICRA research station located at the confluence of the Río Los Amigos and Río Madre de Dios. There he and R. von May, who had been conducting field work in this site for years, observed a thriving population of R. biolat that occupied primarily (though not exclusively) a large bamboo forest. Individuals in this population used the abundant phytotelmata within the bamboo internodes for tadpole rearing. Not surprisingly, the species used other phytotelmata, at CIRCA research station they found tadpoles in tree holes and in a mature floodplain forest in the Tambopata National Reserve, R. von May observed ten tadpoles in fallen bracts of Iriartea deltoidea (one of the most common large palms in western Amazonia). At both localties adults we observed calling outside of the bamboo forests (see Natural History). This behavior was similar to R. lamasi (i.e., from near Tingo Maria). At the CICRA research station and Tambopota National Reserve, R. von May and J.L. Brown encountered individuals that were identical in pattern to lowland populations of R. lamasi, which lacked the characteristic cross on the rostrum (i.e. Fig. 25k). After sequencing several individuals of R. biolat, J.L. Brown found that several individuals were nested within a larger clade containing other R. lamasi (Fig. 3). To reconcile the monophyly of either species (R. biolat and R. sirensis), so that each was a distinct and valid species, numerous populations of R. lamasi would have to be elevated to specific status. In recent years, gene trees have not been congruent with species trees in some studies (particularly mitochondrial gene trees, as used here; see Brown & Twomey 2009 for more info); however, in this case, we have no evidence to suggest that these two species are not simply morphs of a widespread, highly variable species. Within Ranitomeya lamasi sensu Morales (1992), several populations possess unique morphologies (i.e., Fig. 27e, 27k, now classified as morphs 2, 3 and 5 below). Of those sampled in our phylogeny, none of them are reciprocally monophyletic with respect to other morphs (including R. biolat and R. sirensis). Thus, using molecular phylogenetics, behavioral, morphological and

TAXONOMIC REVISION OF RANITOMEYA

acoustic characters, we considered R. biolat and R. lamasi to be junior synonyms of R. sirensis. It seems ironic that the junior synonyms of this species have been subjects of considerable scientific study, though they now bear the name of what once was one of the rarest and most enigmatic poison frogs. Definition and diagnosis. Assigned to the genus Ranitomeya due to the combination of the following characters: adult SVL< 20.0 mm, dorsal coloration conspicuous, dorsolateral stripes extend to top of thighs, brightly colored throat, distinctive pale reticulation on limbs and venter, dorsal skin smooth, finger I greatly reduced and shorter than finger II, finger discs II–IV greatly expanded, disc of finger 2–2.4 times wider than finger width, thenar tubercle conspicuous, toe discs III–V moderately expanded, toe webbing absent, larval vent tube dextral, maxillary and premaxillary teeth absent. Adults use phytotelmata for reproduction and deposit eggs above phytotelm. Tadpole light gray, ovoid, with irregular markings present late in development. Five principal morphs of this species currently are known: (i) the Sira morph (Fig. 26l–n), (ii) the Lamasi morph (Fig. 27e–f), (iii) the Divisoria morph (Fig. 25p–s), (iv) the Biolat morph (Fig. 25j–o) and (v) the Panguana morph (Figs. 25t, 26a–g, j,k, o, p, 27ad). All possess a large, brightly-colored ventral spot (typically yellow) on a bluish to greenish ground color (a possible synapomorphy shared with R. vanzolinii and occasionally R. flavovittata). (i) The Sira morph is an amelanistic morph (lacking any black pigmentation). This morph has a solid red to orange-red dorsum and metallic, sage-green limbs and ventral coloration. The throat and center of the belly each have a large spot, identically colored as dorsum. This is the nominotypical morph for the species and is only known to occur within the Cordillera El Sira, Huánuco, Peru. (ii) The Lamasi morph possesses irregular and broken yellow to yellow-green dorsolateral, ventrolateral and middorsal stripes that often connect to each other irregularly. Venter and limbs are green to blue-green and coarsely reticulated. Most individuals possess bright white to cream spots on the upper surface of the upper thighs, near the groin and upper surface of the forearms, near the axilla. All other morphs have blue to blue-gray reticulation/spotting on venter and limbs on a black ground color. Further, they all possess largely complete middorsal, dorsolateral and ventrolateral stripes. This morph in only known to occur within the lowlands and mid-elevation forests of the Departments Pasco and Huánuco, Peru. (iii) The Divisoria morph has very broad bright yellow dorsolateral stripes (fusing entirely with oblique lateral stripes) and an identically colored, but thinner middorsal stripe. The middorsal stripe fuses with the labial stripe (anterior to the orbits), creating the appearance of a short ‘Y’ or ‘T.’ The middorsal stripe is incomplete (though see below); the stripe breaks in the center of the dorsum, and the black ground coloration between the middorsal and dorsolateral stripes creates a large black “X” that covers most of the dorsum (occasionally this morph is referred to as the ‘X’ morph). Other Divisoria populations are identical in all regards; however, they possess a complete middorsal stripe. The black nostril spots fuse on the rostrum, creating the appearance of an upside-down ‘U’ on the tip of the snout. The venter and limbs are typically light blue to sky blue, giving the appearance of being evenly stippled in identically sized large black spots. This morph is only known to occur within the east versant of the Divisoria range of the southern Cordillera Azul near Aguaytia, Department Ucayali, Peru. (iv) The Biolat morph has thin yellow dorsolateral, middorsal and ventrolateral stripes. A middorsal stripe extends to the tip of the snout, creating the appearance of a crucifix that crosses anterior to orbits. Often the middorsal stripe is slightly darker, appearing yellow-green. The ventrer and limbs are typically blue-green to light gray and can be finely reticulated or evenly stippled in identically sized black spots. This morph occurs predominantly in the lowlands of southern Peru (mostly known from the Department of Madre de Dios and occasionally observed in the Department of Cusco), southwestern Brazil (State: Acre) and northern Bolivia (Department: Pando) where it is typically found in bamboo forests (Maldonado & Reichle 2007; Melo-Sampaio & Barbosa 2009; von May et al. 2009a). (v) The Panguana morph is one of the most variable and widespread morphs. The ventral spot, chin, dorsolateral, ventrolateral and middorsal stripes range from bright red, orange, or green to bright or dull yellow. The middorsal stripe fuses with the labial stripe anterior to the orbits, creating the appearance of a short ‘Y’ or ‘T.’ The black nostril spots typically fuse on the rostrum, creating the appearance of an upside down ‘U’ on the tip of the snout. Most individuals possess large bright white to cream spots on the dorsal surfaces of the limbs at the groin and axilla. The venter and limb coloration is typically bluish green to metallic green to light gray. The limbs and venter can be finely reticulated or have the appearance being evenly stippled in identically sized black spots. This morph occurs within the humid cloud forests and lowland rainforests of Departments Huánuco, Pasco, Junín, Ucayali, southwestern Loreto and southern San Martín. Morphs 2–5 are similar in appearance to R. amazonica, R. toraro sp. nov., R. variabilis, R. ventrimaculata

BROWN ET AL.

and R. imitator; however, all these species lack a large, brightly colored belly spot. No other species of Ranitomeya is easily confused with the Sira morph. Natural history. Recently, von May et al. (2008b) published a detailed description the tadpole of R. sirensis (as Ranitomeya biolat). Our observations of other populations match this description, based on observations on tadpoles of R. sirensis (the Panguana morph) from Puerto Inca and Contamana. In that same publication, von May also noted that this species exhibited male-only parental care. This is supported by other observations, i.e., the lack of observations of trophic eggs, which are commonly observed in pools containing R. imitator and R. vanzolinii larvae (Brown et al. 2008b; Caldwell & de Oliveira 1999). Typically the tadpoles of R. sirensis consumed soft-bodied mosquito larvae (Trichoprosopon digittatum and Culex sp.) and occasionally consumed predaceous mosquito larvae in the genus Toxorhynchites (von May et al. 2009a, b). This species also has considerably larger home ranges than other biparental care species (mean 150 ± 184 m2 versus 31.50 ± 23.05 m2 in R. vanzolinii and 10.9 ± 14.31 m2 in R. imitator; Waldram 2008, Brown et al. 2009b). However, this comparison is not entirely legitimate because Waldram (2008) used a coarse (and unconventional method) to measure space use, sampling data at a spatial resolution of 25 m2, a resolution that frequently exceeds the territory sizes of biparental care species and is almost equal to many male-parental care species (Pröhl 2005; Brown et al. 2009b, Werner et al. 2011). In Tambopata, R. sirensis individuals were observed on average in six grid cells (150 m2/25 m2 = 6). Because of this, it is fair to assume the home ranges for this species are large (even if the individuals only occurred in a small portion of each grid), and are much larger than those observed for any biparental care species. Ranitomeya sirensis has larger clutches than other biparental care species (3.3 versus 1.6 in R. imitator with biparental care and 3.6 in R. variabilis, a male-only parental care species; Brown et al. 2009b). Waldram (2008) suspected that tadpole oophagy was an important food source but never observed egg feeding or trophic eggs after six months of observation. Further, von May et al. (2009a, b) studied this species for 12 months and never observed egg feeding or the presence of unfertilized trophic eggs (which are typically deposited singly or in pairs below the surface of the water). In Tambopata, von May observed the tadpoles of R. sirensis cannibalize the tadpoles of Allobates femoralis, Ameerega trivittata, as well as on mosquito larvae, whenever they cooccurred in the same phytotelm. Because Waldram (2008) did not report whether tadpoles were consuming eggs or newly hatched tadpoles, it is difficult to determine if this species actually consumes embryos, as do, for example, Ecuadorian R. variabilis (Summers 1999) and French Guianan R. amazonica (Poelman & Dicke 2007). Vocalizations. Based on extensive field observations and a small handful of recordings, the call of R. sirensis is a typical vanzolinii-group call. The call is a loud trill, with notes 0.9–2.2 sec in length, repeated at 6–14 notes per minute. Although we have recordings from only a few localities, we have heard calls of this species at every locality included in our phylogeny (including the type locality) and these calls all appear to be very similar. Distribution. This species occurs in Amazonian rainforests of Bolivia (Department: Pando), Brazil (State: Acre) and Peru (Departments: Cusco, Huánuco, Junín, Loreto, Madre de Dios, Pasco, San Martín, Ucayali), Fig. 31. Conservation status. Following the IUCN Red List categories and criteria (IUCN 2010), we tentatively suggest listing this species as Least Concern (LC). Due to the small range of some morphs, e.g., the Sira morph, they may require additional protection.

Ranitomeya vanzolinii Myers 1982 Account authors: J.P. Caldwell, J.L. Brown, E. Twomey, P.R. Melo-Sampaio, M.B. Souza Figs. 3, 4, 9, 23 (j, k), 28, 31 Tables 1, 4–6 Dendrobates quinquevittatus (non Steindachner, 1864)—Silverstone 1975 (partim): p. 33, Fig. 12 (drawing), pattern D and 14 (drawing), pattern E Dendrobates vanzolinii Myers 1982: p. 9, Fig. 4, Table 1 [MZUSP 51597 collected by Paulo E. Vanzolini and Miguel Petrere at Pórto Walter on Rio Juruá, Acre, Brazil];—Vanzolini 1986; Caldwell & Myers 1990: p. 4; Caldwell 1997: p. 211, Fig. 1; Moreira et al. 1997; Caldwell & de Oliveira 1999: p. 565, Table 1–2, Figs. 1–3; Schulte 1999: p. 110, Fig. DB-032, DB-38; Symula et al. 2003: p. 452, Table 1, Figs. 1, 3–6; Christmann 2004: p. 6, Figs. on p. 36. 158; Pröhl 2005: p. 358, Table 1; Roberts et al. 2006a: p. 379, Table 1, Figs. 1, 4; Santos et al. 2009, by implication

TAXONOMIC REVISION OF RANITOMEYA

Ranitomeya vanzolinii—Bauer 1988: p. 1; Grant et al. 2006: p. 171, Fig. 76; Lötters et al. 2007: p. 497, Figs. 498–629; von May et al. 2008a: p. 396, Appendix 2; Souza 2009: p. 31.

Background information. A population of R. vanzolinii from near Porto Walter, State of Acre, Brazil was the focus of intensive behavioral study (Caldwell 1997; Caldwell & de Oliveira 1999) and included in several phylogenetic studies (Symula et al. 2003; Grant et al. 2006). Little information has been published on other populations of this species. Some confusion exists regarding the taxonomic status of R. flavovittata, the sister species to R. vanzolinii (see R. flavovittata account for more details). Tadpole. A total of seven tadpoles of Ranitomeya vanzolinii were collected from Porto Walter, Acre, Brazil, from 4–17 April 1996. A tadpole in Stage 38 (OMNH 36057) was chosen for a complete description. All seven tadpoles were used to examine variation. These tadpoles ranged from stages 26 to 40, although not all stages were represented in the sample. Description of OMNH 36057. Total length 27.8; body length 8.9; tail length 18.9, tail length 67.9% total length. Body depressed, maximum width 6.1, maximum depth 4.4. Snout broadly rounded in dorsal view, rounded in lateral view. Nares small, directed anterolaterally, 0.9 from tip of snout; internarial distance 1.6, interorbital distance 1.5. Distance from naris to anterior edge of eye 1.0, eye width, 1.4. Eyes dorsal, directed dorsolaterally. Spiracle sinistral, short, 1.0, forming small tube free only on outer margin, opening directed posterodorsally 5.7 from tip of snout. Spiracle located 63.8% of body length from tip of snout, just below horizontal midline of body. Vent tube dextral, 0.5 in length. Caudal musculature depth at body–tail junction 3.0; depth of musculature nearly uniform until it begins tapering at posterior one-third of tail. Upper fin originates 1.1 behind tail–body junction, upper fin at midtail 1.0, slightly deeper than lower fin at midtail, 0.8. Tail tip rounded, notochord ends 0.8 from end tail. Oral disc emarginate, located ventrally, not visible from above. Transverse width oral disc 2.4, 38.9% of body width. Lower labium free from body wall. Short, conical papillae present on entire margin of posterior labium and laterally on anterior labium; most of anterior labium free of papillae. Papillae mostly in single row, double in few places. Lower jaw sheath broadly V-shaped, deeper than upper jaw sheath. Transverse width upper jaw sheath 1.1, 44.5% transverse width oral disc. Both upper and lower jaw sheaths with finely serrated cutting edges, serrations becoming smaller on proximal lateral processes of upper sheath, absent on ends of lateral processes. Labial tooth row formula 2(2)/3(1). Labial tooth row A-1 complete, 1.5; A-2 with medial gap, 0.9, right half A2 length, 0.3, 20.0% of A-1. P-1 with medial gap, 0.1; right half P-1 length, 0.5, 44.6% of P-2; P-2 and P-3 equal in length, 1.1. Variation. The two smallest tadpoles, stages 26 and 27, averaged 15.1 ± 1.2 total length. Four tadpoles in stages 34 to 38 averaged 24.8 ± 1.2 total length (range, 22.1–27.8). The single tadpole in early metamorphosis at stage 40 was 24.7; in this specimen tooth rows were becoming irregular, the spiracle opening is beginning to close, and the vent tube is partially absorbed. Tail length as a percentage of total length averaged 62.8 ± 0.01 (range, 60.7–67.9). In the two smallest tadpoles, the upper fin originated at the tail–body junction, whereas in the five largest tadpoles, the fin was low at the tail–body junction, but began to increase in height at an average of 1.4 ± 0.2 (range, 1.0–1.8) behind the tail–body junction. Height of the upper and lower fins was nearly equal in the two smaller tadpoles (upper, 0.40 and 0.45; lower, 0.70 in both individuals). Height of the upper fin was slightly greater in the five larger tadpoles (upper, 1.12 ± 0.10, range 0.90–1.46; lower, 1.01 ± 0.10, range, 0.80–1.38). Tail musculature depth at the tail–body junction was 2.5 ± 0.2 (range, 2.1–2.9) in the five largest tadpoles. Transverse width as a percentage of body width in the seven tadpoles averaged 39.4 ± 0.01 (range, 35.7–45.6). Transverse width of the upper jaw sheath as a percentage of the transverse width of the oral disc averaged 49.7 ± 0.01 (range, 44.5–55.6). A-2 gap is larger than the P-1 gap. Average percentage of the width of the A-2 gap compared to the total width of A-2 is 49.4 ± 0.04 (range, 37.1–60.5). Average percentage of the width of the P-1 gap compared to the total width of P-1 is 17.8 ± 0.04 (range, 8.3–33.6). P-2 and P-3 are nearly identical in length in 5 tadpoles in which both were normal and well-developed. In the smallest tadpole, P-3 was shorter than P-2 but may not have been fully developed. P-3 was deformed in one other tadpole; the ends of P-3 curved upward and fused with P-2. Color in life (based on tadpole observed in a small treehole, later transferred to open-air laboratory). Body dark gray, tail musculature lighter gray, fins opaque light gray. Eye black. First observed 3 March, tadpole nearing metamorphosis by 21 March and had developed small light yellow spots on dorsum.

BROWN ET AL.

Color in preservative. Dorsum of body grayish brown. Venter opaque, intestinal coils dark grayish brown, some organs seen through venter light yellow. Tail musculature light yellow-brown. Back legs grayish brown. Fins translucent. Eyes black. Natural history. Ranitomeya vanzolinii exhibits biparental care and likely, as observed in R. imitator, a monogamous mating system (Caldwell & de Oliveira 1999). After one to two embryos hatch, the male carries the tadpoles on its back and deposits each individually in a phytotelm. Paired males and females remain together in small territories. Males call sporadically each day, and, on average, the female mate responds to his calls every 4.8 days. The male guides the female to their tadpole, they undergo some courtship behavior, and the female deposits two (usually) unfertilized trophic eggs, which the tadpole immediately consumes. The tadpoles are continually fed until metamorphosis. For a more detailed discussion of these behaviors and their evolution see Caldwell (1997) and Caldwell & de Oliveira (1999). In the Reserva Extrativista do Alto Juruá (REAJ) and Parque Nacional da Serra do Divisor (PNSD) of Brazil, this species is frequently observed in primary forests during the rainy season, mainly in areas abundant in bamboo (Guadua sp.), bromeliads, and other phytotelmata (i.e. tree holes; Souza 2009). Adults frequently climb and jump on leaves, stems, and trunks of herbaceous vegetation up to 4 meters in height. In REAJ and PNSD this species uses bamboo internodes, the leaf axils of bromeliads, and other phytotelmata for tadpole deposition (Souza 2009). In forests near Porto Walter, this species uses primarily cavities in small saplings and woody vines for tadpole deposition (Caldwell & de Oliveira 1999). Vocalizations. Like other members of this species group, the call of R. vanzolinii is a loud trill (Fig. 28). Vocalization of Ranitomeya vanzolinii was studied at Porto Walter, Acre, Brazil, using two methods: recordings of males analyzed with Raven 1.3 (Cornell Laboratory of Ornithology, Ithaca, New York), and timing of vocalizations recorded during focal observations of males engaged in distinct behaviors. Focal observations were also made of vocalizing frogs defending territories. Vocalizations of 9 individuals of R. vanzolinii were recorded with a Sony WMD-6 tape recorder. Temperatures at the time of recording ranged from 24.0–27.2ι C. The number of calls analyzed from these recordings per individual varied from 2 to 10 (mean 6.8 calls/individual); calls were averaged within each individual, then across individuals to obtain means for each call variable. The call of R. vanzolinii is a short trill audible from > 5m. Mean number of calls per minute and call duration varied depending on the individual’s behavior (see below; Caldwell & de Oliveira, 1999). Mean number of pulses per call was 16.8 ± 1.7 (range, 8.6–23.8), and number of pulses per second was 31.3 ± 0.8 (range, 27.9–33.6). Dominant frequency ranged from 5162–5412 Hz (n = 7). Some calls revealed four harmonics; the lowest ranged from 2444–2745 Hz (n = 5). During focal observations of males engaging in three behaviors, time of vocalization (to the second) was recorded using a stopwatch. The number of calls per minute as a measure of call rate was obtained during the following behaviors: 1) calling when no female was present; 2) calling when guiding a female to a tadpole rearing chamber; and 3) calling when inside a tadpole rearing chamber with a female. At times when calling from inside a tadpole rearing chamber (i.e., tiny vinehole or treehole), the calls given were much shorter and had only a few pulses, giving the call a drawn-out “creeking” sound. When calling alone, males vocalized at a rate of 3.41 ± 0.25 calls per minute (range, 0–11; n = 6). When calling while guiding a female, males called at a rate of 8.05 ± 0.32 (range, 0–17; n = 8), and when calling while inside a rearing chamber with a female, males vocalized at a rate of 3.13 ± 0.23 (range, 0–16; n = 7). Thus, males called more intensely when guiding a female to a rearing chamber than when calling alone (Mann-Whitney U-test, p < 0.001), or when they were undergoing courtship behavior with a female inside a cavity (Mann-Whitney U-test, p < 0.001). Call rate did not differ when males were alone compared to being in a rearing chamber with a female (Mann-Whitney U-test, p = 0.11). Call duration also varied depending on the male’s behavior. When alone, call duration averaged 0.721 ± 0.08 sec (range, 0.642–0.799; n = 2), when guiding a female 0.378 ± 0.12 (range, 0.256–0.500; n = 2), and when in a tadpole rearing chamber with a female 0.566 ± 0.09 (range, 0.291–0.844; n = 5). Thus, when guiding a female, shorter calls were given at a faster rate. Territories were maintained using vocal bouts in which two resident males in adjacent territories established posts about 2 m apart and alternated calls (Caldwell and de Oliveira, 1999). In one such encounter in which the timing of vocalizations was recorded for 5 minutes, the bout (defined as number of sequential call given by a male before the other male called was 1.23 ± 0.63 calls (range, 1–5 calls). Calls were alternated between the two frogs 74 times during the 5-minute observation period.

TAXONOMIC REVISION OF RANITOMEYA

Distribution. Ranitomeya vanzolinii is a widespread species occurring from premontane cloud forest (approximately 1300 m elevation) in southern Peru, eastward to lowland rainforests in Brazil. This species is frequently observed on the east versant of Cordillera El Sira (where it has been subjected to frequent collection from smugglers for the pet trade) and in western Brazil in the vicinity of Porto Walter, Acre state. The distribution of this species appears to overlap with other members of the vanzolinii group, i.e., R. sirensis in the south of Peru and possibly R. cyanovittata in far eastern central Peru. However to date, there are no localities where two species within the vanzolinii group have been registered. This species occurs in Amazonian rainforests of Brazil (States: Acre and possibly Amazonas) and Peru (Departments: Cusco, Pasco, Ucayali), Fig. 31.

Ranitomeya yavaricola Perez-Peña, Chavez, Brown & Twomey 2010 Account author: J.L. Brown Figs. 3, 4, 9, 23 (c, d), 29. Tables 1, 4–6 Ranitomeya yavaricola Perez-Peña, Chavez, Brown & Twomey 2010: p. 4, Figs. 2–5, 11 [MZUNAP-01-520 (holotype) collected by Pedro Perez-Peña near Lago Preto, 17 km W of Estiron de Ecuador, Provincia Ramon Castilla, Departamento Loreto, Perú, 2009]

Background information. For a summary of current information this species see Perez-Peña et al. 2010. Our phylogenetic results (referenced in Perez-Peña et al. 2010 as unpub. data) place this species in the vanzolinii group, sister to R. cyanovittata. Distribution. This species is known from Amazonian rainforests of Peru in the Rio Yavari region (Department: Loreto) and may also occur in Brazil (States: Amazonas). Fig 29.

Ranitomeya variabilis species group Figs. 3, 4, 9, 14, 32–37 Tables 1, 4–6 A monophyletic assemblage of two species: Ranitomeya variabilis Zimmermann & Zimmermann 1988 and R. amazonica Schulte 1999. Definition and diagnosis. Medium to large adult size (SVL 14–21 mm); two prevalent dorsal patterns: large ovoid black spots on greenish dorsum or complete yellow to red dorsolateral stripes, middorsal stripe and oblique lateral stripes (note that occasionally these species possess ventrolateral stripes, or an intermediate between oblique lateral and ventrolateral stripes, suggesting this trait is a continuous character); large intestine of larvae entirely pigmented; LTRF 2(2)/3(1), oral disc emarginated; larvae gray, tadpoles cannibalistic on embryos and other tadpoles; eggs gray; neither males nor females territorial; promiscuous mating system with male parental care; females produce between 2–6 eggs per mating. Advertisement calls consist of short, regularly spaced buzz calls, notes 0.16–0.44 sec in length, repeated at 24–70 notes per minute (Fig. 14).

Ranitomeya amazonica Schulte 1999 Account authors: E. Twomey, J.L. Brown, M. Pepper, M. Sanchez-Rodriguez Figs. 3, 4, 9, 14, 34 (l–q), 35 (a–q), 36 Tables 1, 4–6, 8 Dendrobates quinquevittatus (non Steindachner 1864)—Silverstone 1975: p. 35 (partim) [French Guiana specimens from Crique Ipoucin, Approuague drainage (LACM 42309); Camopi (MHNP 50-08); Saut Maripa, Fluve Oyapock (MHNP 5009; LG 73); Crique Gabrielle, Mahury drainage (LG 701); Mont Galbao (LG1422)]; Lescure & Bechter 1982: p. 26; Ávila-Pires & Hoogmoed 1997 Dendrobates ventrimaculatus (non Shreve 1935)—Daly et al. 1987: p. 1025 (partim) [specimens from Mishana, Río Nanay drainage, Loreto, Peru]; Kok 2000: p. 19; Estupinan et al. 2001: Lescure & Marty 2001; Fig. 3; Brown et al. 2006 (partim): p. 55, Table 2, Fig. 1; Christmann 2004: p. 6, Figs. on p. 23, 24, 42, 110, 112, 113; Roberts et al. 2006a (partim): p. 381, Table 1, Fig. 3; Noonan & Wray 2006 (partim): p. 1009, Tables 1–2, Figs. 1, 4–5; Poelman & Dicke 2007: p. 217; 2008: p. 271

BROWN ET AL.

Dendrobates amazonicus Schulte 1999: p. 32, Figs. DB-047, PB-045, DB-073, [CRS BD 3 P (holotype) collected by Rainer Schulte at “Bosque UNAP, Iquitos (PERU), ca. 130 m NN.” = Allpahuayo-Mishana National Reserve near Iquitos, Loreto, Peru, ca. 130 m.a.s.l];—Lötters & Vences 2000: p. 250; Symula et al. 2003: p. 453, Table 1; Christmann 2004: p. 17, Figs. on p. 17, 18, 20, 22, 42, 89, 90, 93, 94; Brown et al. 2006: p. 55, Table 2, Fig. 1; Santos et al. 2009, by implication Ranitomeya amazonica—Grant et al. 2006: p. 171, Fig. 76; Lötters et al. 2007: p. 463, Figs. 579–581 Ranitomeya ventrimaculata (non Shreve, 1935)—Grant et al. 2006 (partim): p. 171, Fig. 76; Lötters et al. 2007 (partim): p. 504, Fig. 635; Perez-Peña et al. 2010 (partim): p. 18, Fig. 13

Background information. Specimens of this species have been known at least as early as 1975 from material collected from several localities in French Guiana, when Silverstone (1975) referred to them as Dendrobates quinquevittatus. In the same paper, he mentioned a number of specimens from the Iquitos region, although from his drawings and descriptions all these apparently belong to what is currently referred to as Ranitomeya reticulata. Later, Myers (1982), after having worked with John W. Daly in Allpahuayo–Mishana near Iquitos, suggested that two species of small Dendrobates were present in the area (both of which had been referred to as quinquevittatus in Silverstone 1975). One of these was Dendrobates reticulatus Boulenger 1884 “1883”; however, for the second species, he continued to apply the name Dendrobates quinquevittatus. Caldwell & Myers (1990) began using the name D. ventrimaculatus to refer populations of Dendrobates quinquevittatus sensu Silverstone after the redescription of D. quinquevittatus sensu stricto (now Adelphobates quinquevittatus), including the Allpahuayo–Mishana frogs. Schulte (1999) described the Allpahuayo-Mishana as a new species, D. amazonicus, primarily on the basis of its advertisement call, which was said to be shorter and more highly pulsed than that of D. ventrimaculatus. Schulte also included populations from French Guiana in the definition of amazonicus on the basis of call similarities. This description was criticized by Lötters & Vences (2000), who focused on two issues: (i) the calls of both ventrimaculatus and amazonicus were not described on the basis of quantitative data; and (ii) the name amazonicus was considered a possible junior synonym of D. igneus Melin 1941, described from the same general area. We have dealt with the name D. igneus and concluded that it is a junior synonym of R. reticulata (see account). Studies using genetic data (e.g., Symula et al. 2001; Roberts et al. 2006a; Brown et al. 2006; Noonan & Wray 2006; Twomey & Brown 2008) supported Schulte’s (1999) conclusion that amazonicus was a taxon separate from what was called ventrimaculatus at that time (now considered R. variabilis). Furthermore, putative populations of amazonicus from near Iquitos appeared to be closely related to individuals from French Guiana, further confirming Schulte’s (1999) arrangement. Still, the taxonomic status of amazonicus remained dubious because it rendered ventrimaculatus paraphyletic, yet at the same time there was speculation that ventrimaculatus may actually include several species, so no taxonomic actions dealing with amazonicus were made. One of the main purposes of this monograph is to clarify the taxonomic status of R. ventrimaculata sensu Caldwell & Myers (1990), while simultaneously assessing the validity of R. amazonica. Therefore, since 2004, we densely sampled individuals of both suggested taxa and obtained call recordings from various localities. The results of our phylogenetic and acoustic analyses largely support the conclusions of Schulte (1999). Individuals from near Iquitos and Río Napo in Peru, Leticia in Colombia and throughout French Guiana all form a monophyletic clade with relatively low genetic diversity. This clade is sister to a widespread, diverse clade containing individuals throughout much of the eastern Andean versant and parts of the upper Amazonian basin. The latter contains individuals previously referred to as R. ventrimaculata or R. variabilis (i.e., R. variabilis sensu this paper). Surprisingly, R. amazonica and R. variabilis sensu this paper occur in remarkably close contact south of Iquitos and may even be sympatric in some areas along the lower Río Tigre, Loreto, Peru (M. Sanchez-Rodriguez, pers. comm.). Additionally, call differences do exist between these two species. Ranitomeya amazonica has (on average) significantly fewer pulses per note and a lower pulse rate than R. variabilis (Table 8). However, interspecific overlap occurs in these parameters. As an example, R. variabilis from Cerro Yupati, Colombia and Shucushuyacu, Peru, had notes containing 19.5 and 20.6 pulses (two topotypic R. amazonica had notes containing 20.4 and 21.8 pulses). Also, when measuring pulse rate, one R. variabilis individual from Shucushuyacu had a pulse rate of 106 pulses/ note, while one R. amazonica individual from km. 41 on the Iquitos–Nauta road (Peru) had a pulse rate of 108 pulses/note. We should note that Schulte (1999) described the call of R. amazonica as being shorter and more pulsed than that of R. variabilis sensu this paper. We did not find significant differences in note length, and the pulse-rate data indicate that R. variabilis (rather than R. amazonica) has more highly-pulsed calls (Table 5).

TAXONOMIC REVISION OF RANITOMEYA

FIGURE 32. Ranitomeya Plate 11. variabilis group. A–P: Ranitomeya variabilis (Highland morph): A: Chazuta, San Martin, Peru; B–G: Upper Cainarachi Valley, San Martin, Peru (Ω); H & I: Borja, Loreto, Peru (24, 46Φ); J: Parque Nacional Ichigkat Muja, Amazonas (D. Rodriquez-Mercado); K: Saposoa, San Martin, Peru (Ω); L: Comparision between Saposoa and upper Cainarachi Valley populations (respectively); M: Xanthosoma sp. that was being used by R. variabilis for tadpole and egg deposition near Saposoa, San Martin, Peru; N: Saposoa, San Martin, Peru (Ω); O: Tocache, San Martin, Peru (C. Torres); P: Macas, Morona Santiago, Ecuador (J. Verkade, Ω). (nΦ = number of individual in phylogeny, Ω = population sampled in phylogeny).

BROWN ET AL.

FIGURE 33. Ranitomeya Plate 12. variabilis group: A–P: Ranitomeya variabilis (Lowland morph): A: Contamana, Loreto, Peru (42Φ); B: Shamboyacu, San Martin, Peru (41Φ); C: Lower Huallaga Canyon, San Martin, Peru (‡); D: Callanayacu, San Martin, Peru(‡); E: Barranquita, San Martin, Peru (‡); F: Pongo de Cainarachi, San Martin, Peru (7,8,11Φ); G & H: Bonilla, San Martin, Peru (1,3, 9 Φ); I & J: Varadero, Loreto, Peru (27Φ); K: Quebrada Blanco, Loreto, Peru (Ω); L: Rio Boncuya, Loreto, Peru (G. Gagliardi); M: Upper Rio Nanay, Loreto, Peru; N: Middle Rio Nanay, Loreto, Peru; O & P: Archidona, Napo, Ecuador (EHP and J. Verkade). (nΦ = number of individual in phylogeny, Ω = population sampled in phylogeny). TAXONOMIC REVISION OF RANITOMEYA

FIGURE 34. Ranitomeya Plate 13. variabilis group: A–K: Ranitomeya variabilis: A: Archidona, Napo, Ecuador (J. Verkade); B– D: Macuma, Morona-Santiago, Ecuador (J. Verkade and EHP); E &F: Puyo, Pastaza, Ecuador (j. Verkade) G: Yupati, Vaupés, Colombia (14 Φ); H: Embryos, Archidona, Napo, Ecuador (J. Verkade); I: Breeding pair of R. variabilis, Archidona, Napo, Ecuador (J. Verkade); J & K: Yupati, Vaupés, Colombia (14-16 Φ). L–Q: Ranitomeya amazonica: L: Km 41 Iquitos, Loreto, Peru (ET, 26Φ); M: Km 31 Iquitos, Loreto, Peru (PPP); N & O; Km 26 Iquitos, Loreto, Peru (ET, 25, 27Φ); P: Upper Rio Mazan-Pintuyacu, Loreto, Peru (J. J. Lopez-Rojas); Q: Lower Rio Mazan, Loreto, Peru (J. J. Lopez-Rojas). (nΦ = number of individual in phylogeny).

BROWN ET AL.

FIGURE 35. Ranitomeya Plate 14. variabilis group: A–Q: Ranitomeya amazonica: A–C: Iquitos, Loreto, Peru; D–K: ‘Arena Blanca’, Loreto, Peru (16–18, 20–24Φ); L: 30 km west of Pevas, Loreto Peru; M: Rio Sucusari, Loreto, Peru (B. Pieper); N & O: Nouragues French Guiana (EHP); P: French Guiana (B.P. Noonan, Ω); Q: Estação Científica Ferreira Penna-Caxiuanã, Para, Brazil.(nΦ = number of individual in phylogeny, Ω=population sampled in phylogeny).

TAXONOMIC REVISION OF RANITOMEYA

Dorsal coloration and pattern do not appear to be good characters to distinguish R. variabilis and R. amazonica. Schulte (1999) recognized this by including individuals from French Guiana (which have yellow stripes) under his definition of R. amazonica, even though the populations near Iquitos (around km. 23 on the Iquitos–Nauta road) all have vibrant red or reddish-orange stripes. One of us (E. Twomey, unpub. data) has also found yellow individuals near Iquitos that fall within the amazonica clade, sampled at km. 41 on the Iquitos–Nauta road. Interestingly, individuals sampled from km. 81 (next to Nauta) all fall within the variabilis clade; these individuals are also yellow. In addition to high variation in dorsal coloration in R. amazonica, their patterns have high levels of variation as well. In 2007, a population of R. amazonica was discovered by M. Pepper and M. Sanchez-Rodriguez that was strikingly similar in appearance to R. reticulata (Fig. 35d–k). At the time of discovery, this population was thought to potentially represent a new species; however, after extensive genetic testing (see terminals labeled “Arena Blanca” in the phylogeny), we concluded these individuals were another morph of R. amazonica that may be involved in a mimicry system with the sympatric R. reticulata (see Mimicry section). Biogeographically, R. amazonica represents somewhat of an enigma, although it appears that Schulte (1999) chose the specific epithet wisely, as this species appears to be affiliated with the Amazon River. Noonan and Wray (2006) pointed out that, while R. variabilis sensu this paper appears to be phylogeographically structured along a latitudinal axis, the phylogeographic structuring in R. amazonica appears to be along the Amazon River. Inclusion of additional samples from Río Napo (Peru), Puerto Nariño (Colombia) and Leticia (Colombia) support this hypothesis. It is worth noting that there are large collection gaps between Leticia, Colombia and French Guiana, either explainable with the generally inadequate sampling in this region or natural absence (see Lötters et al. 2010 and discussion). Definition and diagnosis. Assigned to the genus Ranitomeya due to the combination of the following characters: Small adult SVL (16–19 mm), dorsal coloration conspicuous, dorsolateral stripes highly variable and sometimes absent, but when present extend to top of thigh, ventrolateral stripes usually present (usually absent in ‘Arena Blanca’ populations), brightly colored throat, distinctive pale reticulation on limbs and venter, dorsal skin smooth, finger I is greatly reduced and shorter than finger II, finger discs II–IV greatly expanded, disc of finger 2.6 times wider than finger width, thenar tubercle conspicuous, toe discs III–V moderately expanded, toe webbing absent, larval vent tube dextral. Adults use arboreal phytotelmata for reproduction and deposit eggs above phytotelm, maxillary and premaxillary teeth absent. Two morphs are currently known of this species: (i) the Nominotypical morph (Figs. 34 l–q, 35 a–c, n–q), and (ii) the ‘Arena Blanca’ morph (Fig. 35 d–m): (i) The Nominotypical morph has a well-defined black ‘Y’ on the dorsum, with the two ends terminating just medial to the eyelids and the single end terminating on the posterior dorsum. In most individuals, a small black dot is present on the snout (as in Fig. 34m), but in some individuals this spot is absent (e.g., Fig. 35c). The dorsal stripe color is variable. Most populations around Iquitos have dorsal stripe coloration ranging from red to orange, but at least one individual in the area has been found with yellow dorsal stripe coloration. Most individuals have somewhat irregular dorsolateral stripes, and an incomplete middorsal stripe that terminates between the shoulders (i.e., between the ‘fork’ of the black ‘Y’). Ventrolateral and labial stripes are present and typically yellow. Limbs and venter are black with prominent pale reticulation that ranges from bright sky-blue to dark bronze-blue or gray. The throat is brightly colored, usually bright yellow or orange and has irregular black markings around the lateral edges. This morph occurs throughout much of the Amazon basin (see Distribution below). Red and orange individuals have been found from near Iquitos, Loreto, Peru near Río Napo and 70 km N Leticia, Amazonas, Colombia (T. Defler, pers. comm.). Yellow individuals have been found south of Iquitos, in Puerto Nariño, Leticia, throughout much of French Guiana and in extreme eastern Brazil near the Amazon river delta. (ii) The ‘Arena Blanca’ (white sands) morph is a highly variable morph that may be involved in Müllerian mimicry with R. reticulata. This morph generally has a bright orange to red dorsum with irregular black markings. These markings can be manifested as partial stripes (i.e., Fig. 35e) or small spots (e.g., Fig. 35f). In extreme cases, these black markings can be attenuated, rendering the dorsal pattern nearly identical to that of R. reticulata (e.g., Fig, 35k). This morph is known only from a handful of sites in the Department of Loreto, Peru, exact locations undisclosed (available upon request). Tadpole. The description is based on a single tadpole from lower Río Napo, Loreto, Peru. Mouthparts were verified in an individual from an unknown locality in French Guiana.

BROWN ET AL.

FIGURE 36. Known distribution of Ranitomeya amazonica. The inset map displays the geographic extent of distributions (black circles = all other Ranitomeya).

FIGURE 37. Known distribution of Ranitomeya variabilis. The inset map displays the geographic extent of distributions (black circles = all other Ranitomeya).

TAXONOMIC REVISION OF RANITOMEYA

Zootaxa 3083 漏 2011 Magnolia Press 路

Tadpole in stage 29, no external gills, feeding on detritus. Body ovoid in dorsal view, wider near vent. Total length 16.6; body length 9.2; tail length 7.4, 45% of total length. Body width 7.6; body depth 5.3, 70% of body width. Eye well developed; naris small; distance from naris to anterior edge of eye 1.0. Eye positioned dorsally on head, directed dorsolaterally. Spiracle well developed, sinistral; vent tube dextral. Tip of tail bluntly rounded. Tail muscle height at base of tail 2.5; tail muscle width at base of tail 1.9; maximum tail height 3.3. Dorsal and ventral fins approximately of same height. Oral disc ventral, emarginated, transverse width 1.7, 22% of body width. Single row of small papillae present laterally and ventrally; dorsal gap where papillae absent. LTRF 2(2)/3(1) with A-1 developed on upper labium, A-2 with wide medial gap (one-third total width of tooth row); P-1 on lower labium with narrow medial gap; P-2 width equal to P-1; P-3 80% width of P-1. Color in life. Head and body gray, mouthparts visible from above. Abdomen mostly transparent; intestinal coils black; heart visible. Tail musculature uniform gray, dorsal and ventral fins opaque gray. Natural history. The natural history of R. amazonica (mainly French Guiana populations) has been described in detail elsewhere (e.g., Lescure & Bechter 1982; Schulte 1999; Lötters et al. 2007; Poelman & Dicke 2007, 2008); however, we will summarize some pertinent information here. Like other members of the variabilis group, R. amazonica breeds most frequently in bromeliads. Clutches consist of 2–6 eggs are placed in the axils of bromeliads partially submerged in water (Lötters et al. 2007; Poelman & Dicke 2007). Males carry tadpoles to other bromeliad axils and cannibalistic tadpoles are deposited individually. These tadpoles are abandoned (i.e., no trophic egg-feeding) and feed on mosquito larvae, algae, detritus or other tadpoles that may be unwittingly deposited in the same phytotelm. In French Guiana, this species was observed changing its food provisioning strategy. Early in the breeding season, tadpoles subsist predominantly on food within the phytotelm, whereas later in the breeding season (apparently due to increased desiccation risk), parents deliberately deposit embryos as a food source in pools containing their offspring to ensure quick metamorphosis (Poelman & Dicke 2007, 2008). Vocalizations. The call of R. amazonica consists of a series of short buzz notes, note range 0.16–0.36 sec in duration, repeated at a rate of 24–70 notes per minute. Distribution. This species occurs in Amazonian rainforests of Brazil (States: Amapá, Amazonas, Pará), Colombia (Department: Amazonas), Guyana (Region: Upper Takutu-Upper Essequibo), French Guiana (Arrondissements: Cayenne, Saint-Laurent-du-Maroni) and Peru (Department: Loreto), Fig. 36. Conservation Status. Following the IUCN Red List categories and criteria (IUCN 2010), we suggest listing this species as Least Concern (LC).

Ranitomeya variabilis Zimmermann & Zimmermann 1988 Account authors: J.L. Brown, E. Twomey, E.H. Poelman, M. Pepper, M. Sanchez-Rodriguez, P. Perez-Peña Figures 3, 4, 9, 14, 32 (a–p), 33 (a–p), 34 (a–k), 37 Tables 1, 4–6, 8 Dendrobates minutus ventrimaculatus Shreve 1935 (partim): p. 213–214 [paratypes MCZ 19737, 19739 and LACM 44398 (formerly MCZ19741) collected by O.C. Felton in 1933 at “Sarayacu, Ecuador”; paratypes MCZ 19685 and MCZ 19689, collected by C. Spencer between January 1931 and August 1932 “along the Pastaza River from Canelos to the Maranon river, Ecuador”] Dendrobates quinquevittatus (non Steindachner 1864)—Crump 1971; Silverstone 1975 (partim): p. 35; Duellman 1978: p. 125, Plate 1; Lescure & Bechter 1982: p. 26; 1983; Myers 1982: p. 13, Fig. 7; Zimmermann & Zimmermann 1988: p. 132, Schulte 1999: p. 48, Fig. DB-050. Dendrobates ventrimaculatus (non Shreve 1935)—Daly et al. 1987: p. 1025; Schulte 1999: p. 129; Caldwell & Myers 1990: p. 17; Summers & Amos 1997: p. 261; Fujitani et al. 1998: p. 74; Summers 1999: p. 557, Figs. 1–7; Summers & Earn 1999: p. 517; Summers & Symula 2001: p. 17; Symula et al. 2001: p. 2415, Table 1, Figs. 1–3; 2003: p. 453, Figs. 1–6; ; Cisneros-Heredia 2003; Christmann 2004: p. 5, Figs. on p. 15, 42, 98, 100, 147, 151; Brown et al. 2006: p. 55, Table 2, Fig. 1; 2008a: p. 1157; 2008b: p. 2; 2009b: p. 478; Noonan & Wray 2006: p. 1008, Table 2, Figs. 4, 5; Roberts et al. 2006a: p. 378, Table 1, Figs. 1, 3, 4 Dendrobates variabilis Zimmermann & Zimmermann 1988: p. 132 [SMNS 7054 (holotype) collected by Rainer Schulte from "Departamento San Martín, Peru"; restricted to "km 27 auf der Straße von Tarapoto nach Yurimaguas" by Henle 1992: p. 102];—Caldwell & Myers 1990: p. 17; Henle 1992: p. 102; Schulte 1999: p. 39–46, Figs. DB-067, DB-044, DB-063, DB-

BROWN ET AL.

009; Summers & Earn 1999: p. 517; Symula et al. 2001: p. 2415, Table 1, Figs. 1–3, 2003: p. 453, Figs. 1–6; Lötters et al. 2003: p. 1909; Christmann 2004: p. 5, Figs. on p. 25, 121; Noonan & Wray 2006: p. 1008, Table 2, Figs. 4, 5; Roberts et al. 2006a: P. 378, Table 1, Figs. 2–4; Brown et al. 2006: p. 55, Table 2, Fig. 1; 2008a: p. 1140, Table 1; 2008b: p.1, Table 1, Figs. 1–3; 2009b: p. 478, Tables 1–4, Fig. 3; 2009c: p. 148, Fig. 1; Santos et al. 2009, by implication Ranitomeya variabilis—Grant et al. 2006: p. 171, Fig. 76, Fig. 76; Lötters et al. 2007: p. 500, Figs. 630–633; Brown et al. 2008c: p. 9; 2009a: p. 1877, Table 1, 2010: p. 436, Figs. 1, 2, 4, 5; Perez-Peña et al. 2010: p. 18, Fig. 13; Schulte et al. 2010 Ranitomeya ventrimaculata (non Shreve 1935)—Grant et al. 2006: p. 171, Fig. 76; Lötters et al. 2007: p. 504, Figs. 634, 636, 637; Brown et al. 2009a: p. 1877, Table 1; Perez-Peña et al. 2010: p. 18, Fig. 13

Background information. Technically, this species did not receive a formal taxonomic name until Zimmermann & Zimmermann (1988) described a morph of this species as Dendrobates variabilis. Prior to that, most members of this group were referred to as Dendrobates ventrimaculatus or Dendrobates quinquevittatus. Shortly after its description Caldwell & Myers (1990) stated “at present it is not diagnosable from all other populations of D. ventrimaculatus s.l.” and considered Dendrobates variabilis sensu Zimmermann & Zimmermann (1988) as a junior synonym of Dendrobates ventrimaculatus Shreve 1935. Symula et al. (2003) responded to Caldwell & Myers’ synonymy, stating “there are acoustic differences between the advertisement calls of D. variabilis and D. ventrimaculatus… Furthermore, D. variabilis is restricted to the summits of mid-level mountains, whereas D. ventrimaculatus is confined to lowland habitats”. They urged future research, particularly hybridization studies, before a conclusion on the specific status could be reached. Certainly, R. variabilis sensu Zimmermann & Zimmermann (1988) and nearby lowland populations of R. ventrimaculata sensu Schulte (1999) have diverged from each other in some regards (as is apparent in their morphology); however, unlike stated by Symula et al (2003), we are not able to recover any consistent acoustic differences between R. variabilis sensu Zimmermann & Zimmermann (1988) and nearby populations of R. ventrimaculata sensu Schulte (1999); (see R. amazonica account for discussion of acoustic differences between R. variabilis and R. amazonica). Interestingly, we have found evidence of four independent origins of “spotted highland” populations (i.e., morphologically similar to variabilis sensu Zimmermann & Zimmermann 1988) in our molecular phylogenetic analyses. To reconcile the monophyly of either species (R. ventrimaculata sensu Schulte (1999) and R. variabilis sensu Zimmermann & Zimmermann (1988)—using only the type populations), so that each is a distinct species, several additional populations would have to be elevated to specific status. Based on the lack of phylogenetic and acoustic data and consistent morphological or ecological differences, we consider R. variabilis sensu Zimmermann & Zimmermann (1988) to be synonymous with R. ventrimaculata sensu Schulte (1999). Because members of this group were erroneously considered R. ventrimaculata (see R. ventrimaculata acount), the name R. variabilis gains precedence. The evolution of spots seems to coincide with the transition to montane forests from lowland ancestors. The exact cause of this trend is unclear and not all montane populations are spotted; however, generally speaking the trend is prevalent. Additionally, we do not have molecular data for several populations of R. variabilis sensu Zimmermann & Zimmermann (1988). To further complicate the situation, it is common within populations of spotted frogs to observe transitional morphs or an entirely lowland morph. This phenomenon gets to the heart of the problems scientists face when trying to clarify the taxonomy of this group. Under any species criterion, no taxonomy can portray the diversity scientists perceive with complete satisfaction. This species is highly nomadic and appears to be tolerant of suboptimal habitats (in terms of geology, climate and reproductive resources) and is often found in areas unoccupied by any other Ranitomeya (e.g., seasonally flooded forests). This behavior likely reduces the possibility of isolation for periods long enough to result in speciation. Thus, like R. sirensis (sensu this paper), these groups appear to be widely dispersed species with the propensity for local adaptation– though not enough to result in complete speciation. Definition and diagnosis. Assigned to the genus Ranitomeya due to the combination of the following characters: adult SVL < 18.0 mm, dorsal coloration conspicuous, dorsolateral stripes present (in lowland populations), stripes complete when present, extending to top of thigh, brightly colored throat, distinctive pale reticulation on limbs and venter, dorsal skin smooth, finger I greatly reduced and shorter than finger II, finger discs II–IV greatly expanded, disc of finger two times wider than finger width, thenar tubercle conspicuous, toe discs III–V moderately expanded, toe webbing absent, larval vent tube dextral, maxillary and premaxillary teeth absent. Adults use arboreal phytotelmata for reproduction and deposit eggs above or in water. Two main

TAXONOMIC REVISION OF RANITOMEYA

morphs of this species are currently known: (i) the Highland morph (R. variabilis sensu Zimmermann & Zimmermann 1988, Fig. 32a–p) and (ii) the Lowland morph (R. ventrimaculata sensu Schulte 1999, Figs. 33a–p, 34a–k). All morphs possess a large black spot on the snout, anterior to orbits: (i) The Highland morph possesses regular yellow-green reticulation, creating ovoid and large black spots on dorsum. The dorsal coloration often transitions from yellow-green anteriorly to blue-green posteriorly. Venter and limbs are green to blue and typically are evenly stippled with medium to large black spots. The throat is yellow and gular spots are commonly present. This morph is known to occur within rainforests in Ecuador (Province MoronaSantiago) and Peru (Department Amazonas, San Martín, Pasco). (ii) The Lowland morph has thin yellow dorsolateral, middorsal and oblique lateral stripes. The ventrolateral and dorsolateral stripes are typically complete; however, the middorsal stripe typically terminates around the axilla and creates a distinctive black ‘Y’ on the dorsum (Fig. 2g-i). In some populations, dorsal coloration is orange (e.g., near Varadero, Department Loreto, Peru); middorsal stripe is largely complete (e.g., populations near Shamboyacu, Department San Martín, Peru); or dorsolateral and oblique lateral stripes fuse near the axilla (e.g., populations near Pongo de Cainarachi, Department San Martín, Peru; Quebrada Blanco, Department Loreto, Peru; Puyo, Province Morona-Santiago, Ecuador). The ventral and limb coloration is typically blue-green to light gray to blue. The limbs and venter are finely reticulated. This morph occurs predominantly though the lowland forests adjacent to eastern Andean versant, though some populations also occur far from mountains (e.g., those in La Pedrera, Department Vaupés, Colombia; Rio Tahuayo, and Contamana, Department Loreto, Peru). This morph is known from the rainforests of Brazil (State Amazonas), Colombia (Departments Amazonas, Caquetá, Putumayo (tentative), Vaupés), Ecuador (Provinces Napo, Orellana, Pastaza, Sucumbíos) and Peru (Departments Amazonas, Loreto, San Martín, Ucayali). The highland morph is similar in appearance to the local R. imitator and in some populations is involved in a mimicry complex, however, populations of R. imitator that are similar in appearance possess paired nose spots (versus a single nose spot in R. variabilis). For a detailed list of characters see Brown et al. (2008b). The lowland morph is similar in appearance to R. amazonica, R. toraro sp. nov. and some populations of R. imitator, R. ventrimaculata and R. sirensis. Ranitomeya variabilis lacks a large belly spot (present in R. sirensis) and bears a conspicuous black ‘Y’ on the dorsum (versus stripes parallel and extending to vent in R. toraro, R. imitator, R. sirensis and R. uakarii). Distinguishing the lowland morph of R. variabilis and R. amazonica can be difficult if nothing is known about the origin of the individual. Where the species co-exist near Iquitos, R. variabilis typically has thin-yellow dorsal stripes and R. amazonica has broad bright red or orange dorsal stripes (though they can also be yellow). Individuals of R. amazonica from the Guiana Shield are almost identical in appearance to some R. variabilis individuals from Pongo de Cainarachi or Quebrada Blanco in Peru, possessing thin-yellow dorsolateral stripes that fuse with the ventrolateral stripe near the axilla. To confidently identify these populations, genetic or locality data must be used.

TABLE 8. Bioacoustic parameters of R. amazonica and R. variabilis. Sample sizes refer to the number of individuals included. For each individual, 1-5 calls were measured and averaged. Localities included were: R. amazonica: km. 23 IquitosNauta road (7 individuals), ‘Arena Blanca’ (1 individual), French Guiana (1 individual): R. variabilis: km. 81 Iquitos-Nauta road (1 individual), Shucushuyacu (1 individual), Río Tahuayo (2 individuals), Cainarachi Valley (2 individuals), Saposoa (1 individual), Pongo de Cainarachi (1 individual), Cerro Yupati (2 individuals), Shamboyacu (1 individual). Mean recording temperatures were compared to show that there was no systematic difference in recording temperatures between the two species.

species

mean temp. (C)

dom. freq. (Hz)

note length (s)

pulses per note

pulse rate (pulses/note length)

inter note interval (s)

notes per second

amazonica

4930 ± 445

0.261 ± 0.061

27.7 ± 9.8

106.3 ± 16.2

1.44 ± 0.73

0.74 ± 0.20

variabilis

5157 ± 341

0.280 ± 0.113

67.0 ± 43.8

235.2 ± 137.9

1.16 ± 0.46

0.86 ± 0.17

0.913

0.203

0.639

0.02

0.01

0.29

0.15

P-value

BROWN ET AL.

TABLE 9.Table of major phylogenetic groups and corresponding support values. Bremer Support Values (BSV) represents the number of changes for a tree with the particular group and the number without it, posterior probabilities (PP), Bayes Factors (BF), Bootstrap Values (BV). Overall Support was calculated by using the following metric: [(Sum of (BSV values > 10 = 1, <10 = 0; PP values > 95 = 1, <95 = 0; BF values 0-10 = 0, >10 = 1; BS = values > 70 = 1, < 70 = 0) / n of values for which data is reported)*100]. Metric values interpreted are as Overall Support indices: > 90 = high, 90-80 = medium-high, 80-60 = medium, 60-0 = low. Study

Grant et al. 2006

Twomey & Brown 2008

Santos et al. 2009

This study

Overall Support

BSV

PP/BF

BV/PP

mtDNA PP/ nDNA PP/ BF

Index (Value)

Adelphobates

88/1.13

98/100

100/42/*

Med-High (71)

Andinobates

100/*

100/100

100/100/8.52

High (100)

Excidobates

98/0.71

100/100

100/*/*

High (100)

Dendrobates (sensu Grant)

100/1.76

100/100

100/100/*

High (100)

Dendrobates (sensu Silverstone)

100/*

99/100

79/100/*

Med (67)

Minyobates

*/*

98/100

100/*/*

Med (75)

Oophaga

114

100/0.37

100/100

100/99/*

High (100)

Phyllobates

132

*/*

100/100

100/100/*

High (100)

Ranitomeya (sensu this study)

100/*

98/100

100/100/9.96

High (100)

Ranitomeya (sensu Grant et al. 2006)

100/2.41

98/100

100/42/*

Med-High (86)

Test Criterion/ Method of Support Values Genera

Tadpole. The description is based on a tadpole from the Cainarachi Valley, San Martín, Peru. Mouthparts were verified in three individuals from same locality. Tadpole in stage 28; no external gills; feeding on detritus. Body ovoid in dorsal view, wider near mouth. Total length 18.1; body length 8.6; tail length 9.5, 52% of total length. Body width 5.2; body depth 3.1, 60% of body width. Eyes well developed; naris small; distance from naris to anterior edge of eye 1.6. Eye positioned dorsally on head, directed dorsolaterally. Spiracle well developed, sinistral; vent tube dextral. Tip of tail bluntly rounded. Tail muscle height at base of tail 1.8; tail muscle width at base of tail 1.3; maximum tail height 2.0. Dorsal and ventral fins approximately same height. Oral disc ventral, emarginated, transverse width 1.9, 37% of body width. Single row of small papillae present laterally and ventrally; dorsal gap where papillae absent. LTRF 2(2)/3(1) with A-1 developed on upper labium, A-2 with wide medial gap (1/3 total width of A-1); P-1 on lower labium with narrow medial gap; P-2 width equal to P1; P-3 75% width of P-1. Color in life. Head and body gray, mouthparts visible from above. Abdomen mostly transparent; intestinal coils black. Tail musculature uniform gray, dorsal and ventral fins opaque gray. Natural history. Zimmermann & Zimmermann (1984) reported that captive specimens of R. variabilis exhibited biparental care. In contrast, observations in the field indicate that this species has a promiscuous mating system: unrelated tadpoles are commonly deposited within the same phytotelm and females do not return to pools to perform egg-feeding. Oviposition usually takes place at the edge of the water within phytotelmata; clutches contain 3–6 eggs. After 10–15 days, the male transports tadpoles to separate small phytotelmata, where they are deposited individually and thereafter abandoned (Summers 1999; Summers & Amos 1999; Lötters et al. 2007; Brown et al. 2008b). Alternatively, within the same population, clutches can be abandoned shortly after reproduction and all the tadpoles fall into the pool below, often resulting in cannibalism of all but one tadpole (Brown et al. 2008b). The degree of embryo cannibalism varies between populations. Near Tarapoto, Peru, tadpoles were never observed cannibalizing embryos that were deposited in

TAXONOMIC REVISION OF RANITOMEYA

their pools (Brown et al. 2008a, 2009c), whereas in populations from Sucumbíos Province, Ecuador, tadpoles were frequently observed cannibalizing embryos (Summers 1999, Summers & Amos 1999). In both populations, the tadpoles would frequently consume other congeneric tadpoles. This species occupies primary and secondary forests where bromeliads are abundant. It commonly breeds in Guzmania spp. and Aechmia spp. Occasionally this species has been observed breeding in tree holes and in the axils of Heliconia, Dieffenbachia and Xanthosoma plants. Most individuals are transient and in mark-and-recapture studies they demonstrate little site fidelity (Summers & Amos 1999, Brown et al. 2009b). Vocalizations. The call of R. variabilis is a series of short buzz notes, note range 0.14–0.44 sec in length, repeated at 34–62 notes per minute. Distribution. This species occurs in Amazonian rainforests of Brazil (States: Amazonas), Colombia (Departments: Amazonas, Caquetá, Putumayo (tentative), Vaupés), Ecuador (Provinces: Morona-Santiago, Napo, Orellana, Pastaza, Sucumbíos) and Peru (Departments: Amazonas, Loreto, San Martín, Ucayali), Fig. 37. Conservation status. Following the IUCN Red List categories and criteria (IUCN 2010), we suggest listing this species as Least Concern (LC). It has a large distribution and occurs within several protected areas.

Incertae sedis

Ranitomeya rubrocephala Schulte 1999 Account authors: E. Twomey, J.L. Brown, S. Lötters Fig. 38 Table 1 Dendrobates rubrocephalus Schulte 1999: p. 138 [CRS BD 5, “Ceja de Selva, Ostandenabhang und vielleicht vorgelagerte Ostkordilleren, Dep. Pasco und Junín, Höhe vermutlich zwischen 600 und 1500 m” (= Ceja de Selva, eastern Andean versant and perhaps outlaying eastern cordilleras, Departments Pasco and Junin, elevation assumingly between 600 and 1500 m.a.s.l.), unknown collector];—Lötters & Vences 2000: p. 254; Santos et al. 2009, by implication Ranitomeya rubrocephala—Grant et al. 2006: p. 171; Lötters et al. 2007: p. 514

Background information. This suggested species was described on the basis of two specimens, collected likely in 1956 by an unknown collector and deposited at MUSM. Initially four specimens (two lost, two now with the describer) were said to originate from the “Ceja de Selva” (= east-Andean versant) in the Department Pasco, as stated on the original label, which is a more than vague locality. According to the original description, another vague report states that this putative species occurs near Chanchamayo-San Ramón, Department Junín (Schulte 1999: p. 143). Moreover, Schulte suggested that this poison frog may originate from the Cordillera Yanachaga or Cordillera San Matéo. Schulte (1999) described the frog as being small (adult SVL < 15 mm) and entirely black in preservative with a reddish-orange head and throat patch. He speculated that the black color in preservative may be green in life and that R. sirensis (sensu Aichinger 1991) was a related taxon. Additionally, the two specimens available for the original description had two unusual characters: the finger formula was 1 < 4 < 2 < 3, unknown in related small poison frogs which have 1 < 2 < 4 < 3 (according to Schulte 1999), although in our experience fingers 2 and 4 are almost equal in length for most species. Second, Schulte (1999) noted a large central hook on the front of the lower jaw and a respective notch in the upper jaw. Based on Schulte’s description, R. rubrocephala resembles R. benedicta which is known only from parts of the Departments Loreto and San Martín.

FIGURE 38. Type series of Dendrobates rubrocephalus Schulte 1999. A. Holotype (BD 5H, MHNJP 408) and B. Paratype (BD 6P, MHNJP 408).

BROWN ET AL.

Lötters & Vences (2000), despite their apparent skepticism, supported R. rubrocephala as a valid species on the basis of coloration and the unique hand morphology. However, because the original description fails to mention an adequate type locality, because no photos or descriptions exist of live specimens and the two type specimens are so badly degraded (being almost entirely black and badly shriveled with unusual posturing and twisted limbs; Fig. 38), it will be exceedingly difficult to ever confidently ascribe any newly collected specimen as belonging to this species. Further, it is possible that characters described for this species, including the two unusual characters mentioned above, could be artifacts of the poor condition of the specimens. As a consequence, we consider Ranitomeya rubrocephala Schulte 1999 as a nomen dubium.

Discussion Phylogenetics

Typoignore: "andto those of Grant et al. (2006), Twomey & The mitochondrial phylogenetic results of this study are largely similar in Twomey & Brown Brown (2008) and Santos et al. (2009). These studies utilized more gene regions (both nuclear and mitochondrial) and more base-pairs, but contained fewer individuals of Ranitomeya and Andinobates species. In these studies all (2008)" the major groups received considerable support (Table 9) and the relationships among most groups were largely consistent. The placement of Minyobates steyermarki continues to challenge phylogeneticists. In Grant et al. (2006), M. steyermarki was found to be sister to a group containing Andinobates, Adelphobates, Dendrobates, Excidobates, Oophaga and Ranitomeya. In this study and in Twomey & Brown (2008), Minyobates was found to be sister to a clade containing Adelphobates, Dendrobates and Oophaga. We place higher credence in the placement of Minyobates as sister to Adelphobates, as in Twomey & Brown 2008 and Santos et al. (2009), given that the gene regions sequenced in these studies were aimed to address more recent phylogenetic relationships. Further, in the current study, the nodes in the phylogeny adjacent to Minyobates received little support, suggesting the placement of this species was unstable (including the node that would place it sister to Adelphobates). Our analysis of nuclear sequence data resulted in novel phylogenetic arrangements among many dendrobatine genera. When using mitochondrial data, most genera are well supported; however, in a concatenated nuclear phylogeny, the relationships among the genera changed dramatically and almost all relationships among genera add parenthesis received little support. These results suggest that researchers need to consider additional nuclear markers when (2008) dealing with dendrobatine phylogenetics. When single genes are analyzed independently, these relationships vary considerably and, in some instances, they reflect mitochondrial phylogenies, whereas others result in novel arrangements. It appears that these differences do not dramatically influence the results of combined analyses of nuclear and mitochondrial analyses, because topologies are similar to mitochondrial-only analyses. This also could be due to a much stronger signal in the mitochondrial data when compared to the limited variation observed within the nuclear genes. Notes on Taxonomic Changes and the Use of the Name Ranitomeya …the taxonomy of poison-arrow frogs, at the family and generic levels, has been and remains confused and controversial P.A. Silverstone 1975, p. 1 In 2006, shortly after the publication of the taxonomy of Grant et al. (2006), feeling this taxonomy was unnecessary and confusing; we (J.L. Brown and E. Twomey) set out to synonymize the proposed genera for Dendrobates (sensu Silverstone 1975). Grant et al. (2006) subdivided the former Dendrobates into five genera at that time. Due the burgeoning body of literature on species concepts, we were surprised by a paucity of scientific discussion regarding higher-level taxonomy. Like many species concepts, the few generic concepts we were able to find were predominantly theoretical and difficult to implement. Following is a brief summary of many prevalent ideas regarding generic concepts. Mayr & Ashlock (1991) stated “the species category signifies singularity, distinctness and difference, while the higher categories have the function of grouping and ordering by emphasizing affinities among groups of species… (and) a genus (generic taxon) is a monophyletic group composed of one or more spe-

TAXONOMIC REVISION OF RANITOMEYA

cies that are separated from other generic taxa by a decided gap.” Other generic concepts were modified species concepts, such as that proposed by Dubois (1988), stating that evidence for morphological, ecological and genetic discontinuities should be considered, and that taxonomists should use a modified biological species concept and species should be placed in the same group (genus) if able to produce viable hybrids. Winston (1999) stated: Genera are defined not by one character, but by a group of carefully chosen characters. Usually at least some of them are correlated either functionally or genetically and so are present in all members of the groups. The members of the genus do not have to share all characters, however. One or more species in a genus may lack one or more of the diagnostic characters. J.E. Winston 1999, p. 341 Quicke (1993) defined a genus as “one or more species that are believed to be closely related.” In vertebrate taxonomy, the only absolute rule of genera we uncovered was the requirement of common ancestry (Wiley 1981; Dubois 1988; Myer & Ashlock 1991; Quicke 1993; Winston 1999). We advocate that genera represent “natural groups” of closely related species that are easily diagnosable via synapomorphies (morphological, behavioral and genetic). One of us (J.L.Brown) reanalyzed the data of Grant et al. (2006) using Bayesian versus parsimony-based analyses. We generated new molecular data (see Twomey & Brown 2008), thinking this might reveal inconsistencies in the taxonomy proposed by Grant et al. 2006. However, we gradually realized that our frustration with the taxonomy proposed by Grant et al. (2006) was not scientific, but based on personal preferences (only a disdain for the new names applied to the taxa we affectionately called “Dendrobates”). In fact, any argument we proposed to justify the synonymy of Grant et al.’s taxonomy could just as easily be applied in their favor (i.e., monophyly, cohesiveness and diagnosability). More recently, Santos et al. (2009) proposed to synonymize the genera of Grant et al. (2006), stating that: …for the purpose of this paper, some corrections to the Grant et al.’s taxonomy must be made in order to have a nomenclature concordant with well-supported clades, and to avoid ambiguity and subsequent confusion in the literature of this family…Dendrobates sensu lato (including Adelphobates, Dendrobates, Excidobates, Minyobates, Oophaga, and Ranitomeya) was found to be a well-supported monophyletic group (as it was previously); thus the splitting of Dendrobates into several genera is unnecessary (e.g., Excidobates) and Grant et al. did not provide an unambiguous list of synapomorphies for their generic concepts.…. Therefore, we synonymize Adelphobates, Excidobates, Minyobates, Oophaga, and Ranitomeya in the genus Dendrobates. J.C. Santos et al. 2009, p. 44 of supplemental material Like our initial criticisms, we think these criticisms are unnecessarily conservative, incorrect and possibly unscientific. To revert Dendrobates sensu Silverstone (1975) to a single genus (rather than 7 genera) ignores the considerable cohesion that the genera proposed by Grant et al. (2006), Twomey & Brown (2008) and this paper reflect (herein simply referred to as the “Grant et al.’s taxonomy”). Much of this unity has been recognized by dendrobatid taxonomists for the last 40 years, e.g., Silverstone’s (1975) classification of the pumilio group (now Oophaga), or Myers’ quinquevittatus (now Ranitomeya; 1982) and minutus groups (formerly Minyobates minus steyermarki, now Andinobates; Myers 1987). The synonymy of Santos et al. (2009) largely undervalues the existence of morphological and life-history characteristics associated with these genera (see below). Under the more exclusive taxonomy, i.e., Grant et al. (2006), Oophaga is a group of poison frogs that: lacks conspicuous limb reticulation, has an entirely cartilaginous omosternum, has a sacrum fused with vertebra 8, engages cloacal touching during courtship/oviposition, uses small phytotelmata for tadpole deposition, has females as nurse frogs, has males that display a chirp call, has tadpoles that possess a single anterior and posterior larval keratodont rows, and has obligate egg feeding by females. Additionally, Dendrobates (sensu Grant et al.) is a group of relatively larger frogs with conspicuous coloration that lack oblique lateral and ventrolateral stripes, and have smooth dorsal skin texture. Adults are predominanty terrestrial, eggs are deposited terrestrially, and tadpoles are transported, often communally, into larger terrestrial pools. Parental care is done by males and females perform mate guarding. Lastly, Ranitomeya is classified as a group of small Amazonian poison frogs that possess the syna-

BROWN ET AL.

pomorphies of reticulated netting on limbs, greatly expanded finger discs, conspicuous throat coloration and unfused vertebrae 2 and 3.

FIGURE 39. A consensus Bayesian phylogeny based on 2262 base pairs of nuclear DNA from five genes (Rag1, Rhodopsin, Histone H3, 28s, 7th in Absentia). Thickened branches represent nodes with posterior probabilities higher than 85 (those which are not labeled are 100). In addition, some branches of interest with posterior probabilities lower than 85 are labeled.

TAXONOMIC REVISION OF RANITOMEYA

Zootaxa 3083 漏 2011 Magnolia Press 路

FIGURE 40. Consensus of 820 most parsimonious trees for 2622 base pairs of five nuclear genes (Rag1, Rhodopsin, Histone H3, 28s, 7th in Absentia). Thickened branches represent nodes with a consensus percentage of 100%.

94 路 Zootaxa 3083 漏 2011 Magnolia Press

BROWN ET AL.

FIGURE 41. Phylogenetic relationships of genera. Left: Relationships between genera of Grant et al. 2006 based on ca. 6100 bp of nDNA and mtDNA, optimality criterion: maximum parsimony, numbers on branches depict Bremer support values. Center: Relationships between genera of Twomey & Brown 2008 based on 2124 bp of nDNA and mtDNA (Bayesian analysis) and Santos et al.2009 based on 2895 bp of nDNA and mtDNA (Bayesian and maximum likelihood analyses). Number on branches depict: posterior probabilities from Twomey & Brown 2008 / bootstrap values from Santos et al. 2009 / posterior probabilities from Santos et al. 2009. Right: Relationships between genera from this study based on 1011 bp of mtDNA, Bayesian analysis, number on branches depict posterior probabilities. If genus was excluded, focal study lacked corresponding taxa.

The diversity described for the more exclusive genera is on par with the differences observed when comparing Phyllobates to its sister group Dendrobates sensu Silverstone (1975). The continued use of Dendrobates sensu Silverstone inaccurately exaggerates the differences between Phyllobates and all other dendrobatines and undervalues the dramatic diversity within this subfamily. Under the broader definition of Dendrobates sensu Silverstone, the previously mentioned variation is contained within a single genus. The genus Dendrobates sensu Silverstone is most commonly diagnosed by appressed first finger noticeably shorter than second (an exception in Adelphobates quinquevittatus, A. castaneoticus, Excidobates mysteriosus, Minyobates steyermarki, Dendrobates truncatus and Andinobates daleswansoni in which fingers are barely shorter or are equal in length), lack of webbing between toes (however, in Andinobates dorisswansonae and A. daleswansoni toes IV and V are fused), finger discs conspicuous, enlarged in adults to at least twice finger width in fingers II, III and IV (however, in most Andinobates, Excidobates and Minyobates finger discs are moderately expanded and often less than two times finger width) and absence of premaxillary and maxillary teeth (Silverstone 1975; Myers 1982). Santos et al. 2009, stated that “some of the putative synapomorphies for genera used by Grant et al. are ambiguous and others are problematic”, but unfortunately, these authors did not explicitly describe these problems so that we cannot address them specifically. However, these same criticisms can be said about Dendrobates sensu Silverstone (1975). Poison frog systematists have frequently ignored inconsistent characters and typically only report those that agree with the definition of their focal taxon. Further, those characters most commonly used (discussed above) do not provide any obvious diagnostic advantages to those mentioned for the seven genera (in that they are not absolute indicators of the genus/genera). Santos et al. 2009 followed by stating: These include those with sequence errors…,those that are polymorphic intraspecifically (coloration pattern and alkaloid presence), poorly defined (e.g., advertisement call type), or undetermined in

TAXONOMIC REVISION OF RANITOMEYA

most species (e.g., larvae morphology, chromosome number, alkaloid profile, and type of parental care). J.C. Santos et al. 2009, p. 44 of supplemental material Given that all the groups were well supported and were demonstrated to be statistically monophyletic by Twomey & Brown (2008), whose topology is almost identical to their own, no phylogenetic support exists for their justification (i.e., para- or polyphyletic genera). Furthermore, the claim that the presence of “sequencing errors” invalidates these genera is also incorrect, and as mentioned above, their topologies for this group are very similar. We are unsure what support “those that are polymorphic intraspecifically (coloration pattern and alkaloid presence)” provides in favor of Dendrobates sensu Silverstone (1975) versus sensu Grant et al. All dendrobatine genera proposed by Grant et al. (2006) are diagnosable in the absence of alkaloid data. Further, most intraspecific polymorphisms in coloration/pattern actually favor the more exclusive genera of Grant et al. (2006). This is because the latter taxonomy reduces the number of species within a genus and genera are composed of taxa that are more closely related; resulting in the presence of synapomorphies that are absent in the larger group Dendrobates sensu Silverstone. More importantly, the synapomorphies commonly used to diagnose Dendrobates sensu Silverstone are not unambiguous and are problematic (with the exception of presence/absence of premaxillary and maxillary teeth, a character intended by Silverstone (1975) to be used only in conjuction with finger disc size and finger length)(discussed above). Despite their claims that much is “undetermined,” the genera Adelphobates, Excidobates, Minyobates, Oophaga, Ranitomeya, Andinobates and Dendrobates have been the focal subjects of hundreds of studies (over 300 counted at the time of writing, including several books; e.g., see Lötters et al. 2007), and Dendrobatidae is one of the more thoroughly studied anuran families. To claim that exhaustive data on all these species is necessary before new taxonomies can be considered is unnecessarily conservative. Further, when characters are placed in a phylogenetic context, there is considerable phylogenetic affinity for many of the characters proposed (e.g., call data: Erdtmann & Amézquita 2009, Table 5; e.g., behavior, Table 6: see Summers & McKeon 2004, Brown et al. 2009; e.g., tadpoles: Grant et al. 2006, this paper, Table 4), thus, some missing data are not likely to dramatically affect these taxonomies. Lastly the specific characters they mentioned (larval morphology, chromosome number, alkaloid profile and type of parental care) were never listed as explicit synapomorphies by Grant et al. (2006) for dendrobatine genera, but rather were included as known characteristics of these genera in addition to the unambiguous synapomorphies. Recently, several systematists have argued that taxonomic changes affecting genera should be minimized to maintain communication among the biological community and in place of creating new genera, subgenera (or unranked taxa) should be created to formally recognize diversity to facilitate nomenclatural continuity (D.C. Cannatella, pers. comm.). We agree that researchers should refrain from needlessly creating genera. The creation of subgroups to maintain consistency, however, downplays a group’s cohesiveness (when compared to the creation of a new genus). This also entrenches taxonomists in previously described genera and can greatly attenuate the biological unity observed in taxonomies resulting from detailed molecular taxonomies (many times replacing genera prescribed from purely morphological taxonomies). Lastly from the point of view of conservation, it is important for all non-specialists to realize how diverse a taxonomic group actually is, rather than masking the diversity within described genera. Ultimately, the current discussion regarding generic classification has little to do with taxonomic code, and more to do with the tradeoff between nomenclatural consistency on one side and adequately depicting a group’s cohesiveness on the other, and hence is open to debates among the scientists. In the case of Ranitomeya, and likely in other dendrobatine genera, it is impossible to maintain taxonomic continuity because of numerous alpha-taxonomic problems, genus-level debates aside. We think it is important that taxonomies be challenged and exhaustively discussed; we hope for generations to come taxonomists will continue to feverishly work to perfect taxonomies. Further, no taxonomic change should be apathetically accepted: continued dialogue (e.g., that of Santos et al. 2009) and careful attention to details will ultimately reduce unnecessary confusion from nomenclatural irresponsibility. Regarding the use of Dendrobates sensu Silverstone (1975), we feel it is time move forward, and we have presented our case in this paper.

BROWN ET AL.

FIGURE 42. Ranitomeya tadpole morphology. Major features of: A. Oral disc morphology, B. Body coloration, C. Body morphology.

Tadpole Morphology Relatively few tadpoles of species of Andinobates and Ranitomeya have been thoroughly described. Among the 28 species, tadpoles of five species are unknown (Table 4). Among the known tadpoles, some descriptions are incomplete and others are based on only one tadpole or only on back-riding tadpoles. Descriptions based on a series of specimens are available for only two species. Back-riding tadpoles that are collected from an adult frog may not have all characters developed. In particular, very small back-riding tadpoles may not have developed labial tooth rows or they may have developed only one anterior and two posterior rows. Even small tadpoles in stages 25 or 26

TAXONOMIC REVISION OF RANITOMEYA

Zootaxa 3083 漏 2011 Magnolia Press 路

may not have rows fully developed. Upper and lower tail fins may change in height or shape in older tadpoles, and proportions of various characters may change as tadpoles develop. The lack of sufficient tadpole specimens for description is a result of several factors. Species in these two genera typically have small clutch sizes, and tadpoles develop in small phytotelmata, such as bromeliads, axils of various plant species, and vineholes or treeholes. Thus, collecting a reasonable series of tadpoles of any species at one time can be difficult. Tadpoles are more easily located for those species that use axils of plants such as Heliconia that occur in stands. Other species use widely spaced bromeliads or treeholes, and their tadpoles, which often occur singly, can only be located by following adult frogs when they visit tadpole nurseries to transport a tadpole or to provide nutritive eggs, or by locating and searching phytotelmata for tadpoles. In addition, relative few researchers work with tadpoles. Understandably, many researchers are interested in adult frogs, which may be poorly known and are necessary to understand relationships and construct phylogenies. Coupled with the fact that many of these species have very restricted ranges often in remote areas means that locating tadpoles is not the highest priority. However, good tadpole descriptions can add to our knowledge of the species and provide valuable clues to relationships (Caldwell et al. 2002). In the future, we encourage investigators to preserve good series of tadpoles from as many localities as possible. We realize that in this paper most of the new tadpole descriptions used only a couple of specimens from a single locality. In this case, our attention (J.L. Brown) to collecting complete tadpole series, regrettably, actualized shortly after completing most the field work for this project. Ideally, specimens in a variety of stages of development will add to the completeness of the description. Only with a good series of well-preserved tadpoles (preserved and permanently stored in 10% formalin) can investigators provide complete descriptions, assess variation, and document how morphological changes occur during development.

Müllerian Mimicry Mimicry appears to be more common than previously documented among Ranitomeya species (Symula et al. 2001). The striking similarities among populations of Ranitomeya uakarii (which possess yellow dorsolateral and middorsal stripes) and Ranitomeya toraro sp. nov., some syntopic populations R. amazonica and R. reticulata, and Varadero populations of R. imitator, R. fantastica and R. variabilis are similar to those in the Müllerian mimicry systems observed by Symula et al. (2001). Assuming the aforementioned cases are in fact a result of Müllerian mimicry, this would mean half of all extant Ranitomeya species are involved in mimicry systems (Fig. 43). In other species, it is apparent that other factors affecting phenotype do not fit a mimicry model. For instance, the reoccurring origins of spotting in montane populations of R. variabilis or the considerable variation observed in R. sirensis suggest that other factors (i.e., sexual selection, genetic linkage, or stochastic processes) played a key role in the evolution of phenotypic diversity.

Dendrobatid Frogs and the Pet Trade Smuggling for the pet trade remains a serious threat to native populations, including Ranitomeya. Despite having brought attention to this (Gorzula 1996; Pepper et al. 2007; Lötters et al. 2008; Brown et al. 2008c; Nijman & Shepard 2010), the pet trade continues to drive the market for smuggled animals (Table 10). For example, shortly after describing R. benedicta (Brown et al. 2008c), publishing new pictures of new morphs of R. imitator (ca. 2006 on www.dendrobates.org), or publishing the rediscovery of E. captivus (Twomey & Brown 2008), the result was the almost immediate arrival of illegal specimens in trade shows in North America and Europe. These acts appear to have had negative impacts on at least some populations. For example, during recent trips to Shucushuyacu (in 2008 and 2010), the type locality of R. benedicta, we were faced with the dismal reality that locals were cutting trees to access the large Aechmea bromeliads growing in the canopies in search of this species. If the hobby is to have a sustainable future, these trends need to cease and hobbyists need to make educated and ethical decisions when purchasing new frogs. Today, consumers have an unprecedented number and variety of sustainably produced frogs available. These choices include captive-bred specimens (particularly those whose founding stock are of legal origin), wild caught individuals imported legally via controlled quotas, and those frogs sustainably bred in the country of origin.

BROWN ET AL.

FIGURE 43. Potential examples of mimicry in Ranitomeya. A. Ranitomeya uakarii and R. toraro from western Brazil. B. Ranitomeya amazonica and R. reticulata from undisclosed locality in Loreto, Peru. C. Ranitomeya imitator, R. variabilis, R. fantastica from Varadero, Loreto, Peru. In this population, individuals of R. imitator display a cline of morphologies, usually closely resembling local morphs of either R. variabilis or R. fantastica. D. Ranitomeya imitator, R. variabilis and R. fantastica from Pongo de Cainarachi, San Martin, Peru. In this population (and surrounding areas), individuals of R. imitator and R. fantastica (nearby populations possess considerably more orange on the head and dorsum and lack the appearance of 2 black stripes down the dorsum) appear similar to widespread lowland morph of R. variabilis. E. Ranitomeya imitator and R. variabilis from Cainarachi Valley, San Martin, Peru. F. Ranitomeya imitator and R. summersi from Sauce, San Martin, Peru. See Symula et al. 2001 for analyses of D, E and F as cases of Müllerian mimicry.

One aspect of amphibian smuggling that is largely overlooked is the process by which most frogs arrive to the market. Simply put, the number of individuals for sale does not reflect the number of individuals removed from the wild. Most illegal specimens are exposed to horrendous shipping and living conditions on the way to the market (Fig. 44). At trade shows only the few specimens that arrived in fair physical condition are encountered. Only the smugglers know the exact mortality rates and number of frogs that are illegally harvested. Based on observations of confiscated shipments, it is apparent most smugglers give little consideration to maximizing the survival of each specimen. Frogs are often packed inhumanely in crowded containers or cages and often food-deprived for days (INRENA personal comm. 2005). Because new frogs (of known illegal origin) continue to arrive at trade shows, many thousands of individuals likely die worldwide each year. For example, in Peru hundreds to thousands of animals are collected by local farmers and held for a middleman until numbers are sufficient to warrant travel costs (Pepper et al. 2007). A smuggled frog can be sold for over 500 times their initial investment (for example in 2008, the first specimens of R. benedicta were sold for over $500 each in Europe). This recipe makes exploitation not only affordable for the smugglers but also so potentially lucrative that there is little incentive to “invest” in proper care of frogs during transit.

TAXONOMIC REVISION OF RANITOMEYA

100 路 Zootaxa 3083 漏 2011 Magnolia Press

BROWN ET AL.

FIGURE 44. Illegal traffic of poison frogs. A. A cage for poison frogs containing ca. 50 A. silverstonei awaiting collection for illegal trade (near Tingo Maria, June 2007) B. Inside view of cage. C. 500 ml soda bottles and D. wooden boxes containing dozens of R. imitator, R. variabilis and R. fantastica awaiting collecting for illegal trade (near Varadero, Loreto, August 2007). E. Dead Ranitomeya vanzolini and Excidobates mysteriosus (an endangered species) found hidden amongst a shipment of tropical fish and destined for a major distributor of frogs in Europe in 2005. Over 400 frogs were discovered in this shipment, concealed among tropical fish exports from Iquitos, Loreto, Peru. The frogs themselves were packed against each other, with no room to move, inside plastic tubes with a diameter of about 3 cm. Nearly half of the animals were dead upon seizure in Lima. The remaining live animals were immediately released to the custody of Zoocriadero Exotics Frogs in Iquitos, and despite considerable effort, more than 80% of the survivors died shortly thereafter. Assuming many more perished during the collection of these animals and transport to Iquitos (as both species habitats occur greater than 600 km from Iquitos), it is easy to speculate that several hundred additional animals perished during this single export. The shipping conditions observed here are standard practice in the illegal trade of poison frogs.

TABLE 10: Summary of information known regarding the current state of smuggling poison frogs for the pet trade. Number – Species

Case Info

Mortality

Punishment

Citation

580 - O. pumilio 22 - D. auratus

In 2004 CITES officials discovered over 600 frogs packed into more than 200 film canisters. Each canister contained 2 to 3 frogs. Three Belgian citizens, arriving from Panama, were arrested.

10%

Arrested, punishment not disclosed

TRAFFIC 2011

196 - O. histrionica

In 2000 a Panamanian man was arrested in Bogota, Colombia transporting Harlequin poison frogs and 344 Matamata Turtles (Chelus fimbriatus). (Destination unspecified, presumed Panama)

Unknown

2 years imprisonment and fined $13,864 USD

TRAFFIC 2011

770 - D. leucomelas

In 2000, Frankfurt, Germany customs officials seized 770 poison frogs hidden in three plastic containers hidden within the hand luggage of the person traveling from Venezuela.

Unknown

TRAFFIC 2011

32 - adults of the following species: D. tinctorius, R. amazonica, Allobates femoralis

In 2006, Belgian federal police stopped 7 Belgians and a French person returning from French Guiana. The latter had the 32 frogs in film canisters in the bottom half of a drink thermos. Belgian police searched their residences and confiscated 10 additional ‘Dendrobates’ specimens.

Unknown

Two perpetrators were fined 2000 Euros each, another 800 Euros. In addition, the three were fined a joint sum of 3000 Euros.

TRAFFIC 2011

373 - of the following species: O. pumilio and D. auratus

Frogs were found in plastic tubes in the suit case of a British passenger arriving from Panama. Resulted in raids on several premises in southern England; additional poison frogs were seized

Unknown

TRAFFIC 2011

330 - O. pumilio 30 - D. auratus

Collected in Costa Rica, a German citizen attempted to hide the specimens in a concealed compartment of a sport bag. Frogs were contained in three plastic boxes.

50% were found dead on arrival, or died since arriving.

Unknown

TRAFFIC 2011

279 - A. galactonotus 281 - D. tinctorious

In 1999, the 560 frogs destined for Germany were confiscated in Guarulhos International Airport, Brazil packed individually in film canisters inside of two suitcases.

Unknown

Postoni & FelipeToledo (2010)

400 of the following species: E. mysteriosus and R. vanzolinii

In 2005 in Lima, Peru officials discovered 400 frogs pack tightly inside of small plastic tubes (ca. 3cm) hidden amongst a shipment of tropical fish arriving from Iquitos intended for Europe.

50% were dead upon discovery, more than 90% eventually died.

Unknown

M. Pepper, M. SanchezRodriguez, unpub. data

TAXONOMIC REVISION OF RANITOMEYA

101

Below we offer some anecdotal reports which we hope give insight to the ramifications of illegal smuggling practices.

Atalaya, Ucayali, Peru A small isolated town on the east versant of the Cordillera El Sira, (Atalaya, Peru) was often visited by foreigners in search of R. vanzolinii. In 2007, we (M. Pepper) arrived here and before leaving the airport, we were approached by an airport security worker with an offer to purchase this species. This offer was made without solicitation or any indication that we had come looking for the frog. After three days of searching near Atalaya, we failed to find a single individual of R. vanzolinii. Subsequent trips to Atalaya in 2007 and 2008 (by M. Sanchez-Rodriguez), as well Shepahua, where R. vanzolinii was also once common, resulted in only one animal being found. Around that time, large numbers of this species arrived in Europe and the US, despite neither Peru nor Brazil ever having issued CITES export permits for the species for non-scientific purposes.

Tingo Maria, Huánuco, Peru Local farmers from the mountains east of Tingo Maria (locally known as the “Divisoria”) recount the times in the late nineties when they would sell frogs to certain “gringos.” One villager claimed that 900 R. sirensis (at the time, known as lamasi) were sold on a single occasion. It is impossible to determine exactly how many frogs were collected for the pet trade; however, it is evident that this unique morph, known only from this area (the Divisoria morph), once was frequently available in the pet trade. Despite frequent visits and arduous searches by several of the authors over the last five years (J.L. Brown, E. Twomey, M. Pepper, M. Sanchez-Rodriguez), we have failed to locate this morph in the wild. Another victim is Ameerega silverstonei, also part of this trade circuit. This species continues to be collected for illegal exportation —one farmer stated that he continues to sell “40–60 adult A. silverstonei a month” and has done so for the last several years (name undisclosed, pers. comm.; Fig. 44). Despite this, relatively few A. silverstonei have arrived on the black markets, suggesting to us that most of them probably died during transit through the sweltering lowland cities of Iquitos and Pucallpa. Unfortunately, these practices are not restricted to Atalaya or Tingo Maria. We have encountered similar situations in dozens of villages and towns throughout our travels in Peru (e.g., Chazuta, Pongo de Cainarachi, Iquitos, Tarapoto, Shucushuyacu and Puerto Inca). While doing the final revisions of this monograph, we decided to remove several localities and pictures because we feared that this information would be exploited by smugglers. We realize that this is not standard procedure; however, people who work with or study marketable species must consider the potential impact on wild populations resulting from novel information published (both in manuscripts and on the internet). These actions detract from the completeness of this research, in essence slowing scientific progress by inhibiting the dissemination of information. It should be noted that the authors are not opposed to keeping poison frogs as pets, in fact the contrary. We feel that dendrobatids play an important role in captivity. They serve as ambassadors to not only the plight of the world’s amphibians but also, more importantly, to the plight of the tropical forests and other natural wonderlands across the globe (e.g., Madagascar). While in captivity, these frogs have become for many a tangible connection to tropical rainforests in far off places. For the many that will not be as fortunate as we are, to have had the opportunity to search amidst the rainforest for these frogs, the only contact they may have with these remarkable forests and their inhabitants are through the exposure to animals in captivity.

Future Directions Ironically, one of the biggest factors impeding anuran taxonomy in South America is the arduous and often long process to obtain collection and export permits (both for whole specimens and genetic samples). Without increased governmental support, rigorous taxonomic studies containing large type series (distributed at multiple museums) are not possible (for both national and international researchers). Recent restrictions on the export of genetic sam-

BROWN ET AL.

ples, with particular fear of “biopiracy” have further complicated this process. It is impossible to plan out effective conservation strategies without basic knowledge of biodiversity. In recent decades, this has become particularly important; we are now witnessing unnaturally high extinction rates (Barnosky et al. 2011). We are among the first generation of scientists to be commonly confronted with the challenge of describing species as they disappear. This problem is further exacerbated by the decreasing emphasis being placed on taxonomy as an important science. In recent decades, funding for collection and taxonomic training has been greatly reduced (both in developed and developing nations), despite wide acknowledgement that quantifying biodiversity is one of the most fundamental requirements for effective conservation (McClain 2011). In regard to the taxonomy of Ranitomeya, we still lack basic life history information for several species (e.g., mating system, reproductive resource types, tadpole development and ecology). Though this information is partially known for most species and, at minimum, for species groups, detailed knowledge of this information can be very important in context of evolutionary studies. Numerous geographic areas remain underexplored or lack genetic material and whole specimens, including adults and especially tadpoles. We have classified three key areas in need of basic exploration based on current collection data and ecological niche modeling: 1. Amazonian Colombia northwest of La Pedrera. 2. Western Brazil (specifically the Rio Jurua and Purus drainages south of Rio Amazonas and the Rio Negro drainage to the north) 3. Southeastern Venezuela (Roraima region and central Guyana). These areas may or may not contain Ranitomeya species; however, they appear to be climatically suitable and are geographically proximate to known populations (see Lötters et al. 2010). In many areas, even if we understand the composition of the Ranitomeya communities, we lack crucial genetic material (see Fig. 45). We hope this revision has revealed some of the problems poison frog taxonomists encounter when delimiting species solely on morphological characters. Still, when using molecular data, researchers also need to proceed with caution, as these tools are not absolute barometers of a species. Factors such as historical introgressive hybridization, incomplete lineage sorting, recombination and gene duplication are increasingly reported (Brown & Twomey 2008, Shimada et al. 2010; Hauswaldt et al. 2010). These phenomena make it difficult to discern a “true” species tree from a gene tree. Arbitrary genetic metrics such as branch lengths or genetic distances (which only vaguely depict common ancestry) can be difficult to interpret and can be misleading. These metrics can be sensitive to the breadth of genetic sampling, demographic processes (i.e., population size and dispersal potential), and the biogeography of the species. Often high genetic divergence can be explained simply by isolation-by-distance rather than speciation. For example, in this study we observed high genetic divergence within R. variabilis; however, when we classify the morphological divergence within this group in context of the mitochondrial phylogeny, we are not presented with any natural subgroups that display consistent differences in morphology —both adult and larval — acoustics, ecology or behavior. In contrast, the R. fantastica species complex (sensu Brown et al. 2008c) occupies a considerably smaller geographic range and demonstrates about a quarter of the genetic divergence observed in R. variabilis. However, the three recognized species within this group differ acoustically, morphologically and phylogenetically. If we were to use the observed genetic divergence of the fantastica species complex as the rule for genetic divergence per species in Ranitomeya, this would suggest that the R. variabilis is a complex containing multiple species, rather than being a single, widespread species. In conclusion, we urge future taxonomists to consider behavioral, morphological (both adult and larval), ecological and phylogenetic data when classifying new dendrobatine species. The inclusion and careful consideration of these data can reduce unnecessary taxonomic confusion and will facilitate the designation of other new species. This revision resulted in the synonymy of five species (R. ignea, R. duellmani, R. biolat, R. lamasi and R. intermedia), description of one new species (R. toraro), redescriptions of three species (R. uakarii, R. variabilis and R. sirensis), a restriction of Ranitomeya resulting in the description of one new genus (Andinobates) and placement of one species as nomen dubium.

TAXONOMIC REVISION OF RANITOMEYA

103

FIGURE 45. Geospatial distribution of our current sampling of Ranitomeya. A. Distribution of Ranitomeya species (each color represents a different species). In figures A–F; warmer colors represent high values. B. Species richness. Spatial density of observations: C. reported localities and D. available genetic material. E. Species distribution model of the genus Ranitomeya. Higher values (represented by warmer colors) depict areas of high habitat suitability. F. Key areas in need of future research (red= areas in need of genetic exploration, yellow= areas in need of basic exploration). Google Earth files downloadable from: http://www.jasonleebrown.org/Ranitomeya_GoogleEarth.zip

A-E;

BROWN ET AL.

Acknowledgments We are grateful to Roy W. McDiarmid (USNM), Addison Wynn (USNM), Jose Rosado (MCZ), Neftali Camacho (LANHM), Jim McGuire (MVZ) and Dan Portik (MVZ) for assistance with collections; David Wake for AmphibiaWeb and Darrel R. Frost for Amphibian Species of the World; Ted Kahn for drawings in Figure 3; Karen Siu Ting, Jesús H. Córdova, Claudia P. Torres, César Aguilar for continued support and assistance at MUSM and DGFFS (formerly INRENA); we thank Karina Ramirez and Rosario Acero Villanes of DGFFS for assistance with the process of obtaining research, collecting and export permits; Isaías J. Reis of ICMBio for assistance with the process of obtaining collecting permits. Karen Fieselman, Bryan Richardson, Jacqueline Rivers-Pepper, Wouter Olthof, and Justin Yeager for assistance in the field; Biology Departments at Duke University and ECU, Trip Lamb and Anne Yoder for support and guidance; Lauren Chan for helping with cluster computing at Duke University; to the students of the Universidad de Los Andes, Bogotá, particularly Sandra Marcela-Cely, Iliana Medina, Osvaldo GilG., Erika Nathalia-Salazar, José D. Gil-Acero, Adriana Dorado, Beatriz Eugenia-Velasquez, Catalina Gonzalez, Giovanni A. Chaves-Portilla, Diana M. Galindo-Uribe, Luis Alberto ‘Beto’ Rueda and Fernando Vargas-Salinas, for their assistance and friendship while in Colombia. We thank Mathieu Chouteau for translating the abstract into French. We are indebted to J. C. Santos for graciously providing several unpublished sequences for use in this project. J.P. Caldwell especially thanks V. de Oliveira for her help in studying R. vanzolinii in Acre, Brazil, A. Lima for arranging logistics for fieldwork near Manaus, Brazil, T.C. Avila-Pires for arranging logistics in Amazonas and Acre, Brazil, and L.J. Vitt for contributions to all aspects of Amazonian fieldwork. D. Mejía-Vargas thanks Dendrobatidae Nederland for their continued assistance. We are indebted to Mago Medina-Müller, German Chavez, Víctor R. Morales, William E. Duellman, Pedro Ivo-Simões, M. Ramírez Zárate and his colleagues at CIMA (Investigación y Manejo de Areas Naturales: Cordillera Azul), Sergio Moreno, Juan Carlos Suarez, Mathieu Chouteau, Wouter Olthof, Manuel Miranda, Manuel Guerra Panaifo, Manuel Fashe, César Lopez Dominguez and Valentina Express, Rainer Schulte, Roy Santa Cruz, Luis A. Coloma, Mathieu Woldhuis, Jan Verkade, Tim Green, Julian Mejía-Vargas, Marcus Bartelds, Pascual Tafur-Gonzalez, César Barrio-Amorós, Juan Carlos Santos, Giuseppe Gagliardi, Tim Paine, Tiffany Kosch, David C. Cannatella, Taran Grant, Brice P. Noonan, Thomas Ostrowski, Karl-Heinz Jungfer, Mathias Kneller, Heike Pröhl, Miguel Vences, Steffen Reichle, Pablo Venegas, Lisa Schulte, Philine Werner, and Steven Chen for sharing their knowledge on poison frogs; José Vicente Rodríguez-Mahecha and Conservation International for logistical support while in Colombia; the efforts of the Sira Teams (not mentioned in text): Bernard Bettens, Charles Nishihira, Tim Paine and Brad Wilson. We thank the following people for their photo contributions: Brad Wilson, Justin Yeager, Thomas Ostrowski, German Chavez, Karl-Heinz Jungfer, Julian Mejía-Vargas, Albertina P. Lima, Walter Hödl, Victor Mora-Luna, Giovanni Chaves-Portilla, Pedro Ivo-Simões, Santiago Ron, A. Blasco Z., Mario Callegari, J. M. Rejifo, G. Gagliardi, A. Toebe, Guillermo Knell, Jan Stenicka, D. Rodriguez-Mercado, Jan Verkade, Bernd Pieper, Brice P. Noonan, Dennis Nilsson, and J. Lopez-Rojas. Research in Brazil was supported by National Science Foundation grants (DEB-9200779 and DEB-9505518) to L. Vitt and J.P. Caldwell and support from SUFRAMA to P. R. Melo-Sampaio. Travel to Peru was funded by NSF (IOB-0544010 and DEB-0134191), National Geographic Society (7658-04) and East Carolina University Research and Development Grant (2006) as part of larger projects. We thank Brian Sexton who generously donated money which helped finance our Contamana expeditions. J.L. Brown was supported during the later stages of this project by NSF DEB 0516276 and a NSF Bioinformatics Fellowship 0905905. Research permits were obtained from the Ministry of Agriculture (INRENA), now called Direcion General Foresal y Fauna Silvestre (DGFFS), in Lima, Peru (Authorization N° 050-2006-INRENA-IFFS-DCB, N°. 0672007-INRENA-IFFS-DCB, N° 002765-AG-INRENA, N° 061-2003-INRENA-IFFS-DCB, N° 050-2006-INRENAIFFS-DCB) and the Ministry of Natural Resources (INRENA) in Iquitos, Peru (Authorization Nº 320-2009AG-DGFFS-DGEFFS), ICMBio in Brazil (Authorization Nº 19347-1). Permits (for JPC) to conduct research and collect specimens in Brazil were issued by Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq, Portaria MCT 170) and the Instituto Brasileiro do Meio Ambiente e dos Recursos Naturais Renováveis (IBAMA, 073/94-DIFAS), respectively.

TAXONOMIC REVISION OF RANITOMEYA

105

DNA sequences: Incertae sedis The following list represents DNA sequences of dubious origins (both in terms of tissue source or the locality applied to the tissue source) or those that differ considerably from other individuals from similar populations (whereas other gene regions sequenced share considerable similarity). In some cases we cannot rule out that these sequences are not the “true” sequences; however, if used in the future, results should be interpreted with caution. 1. DQ502302.1: Phyllobates lugubris isolate 329 histone H3 gene. Matches (100%) a caecilian (Caecilia tentaculata) histone H3a gene and is not similar to any Phyllobates. 2. DQ371305.1: Dendrobates uakarii (now considered R. uakarii) 12S ribosomal RNA gene. This sequence is very similar to several individuals of Adelphobates castaneoticus individuals rather than other Ranitomeya. Since this realization, this individual has been resequenced and the new, correct 12S sequence is available (see Apendix II). This error can be definitively attributed to a “tracking error” during post DNA sequencing processing. This explains the extremely long branch lengths observed in Roberts et al. (2006a) and Brown et al. (2006). 3. AF482773.1: Dendrobates quinquevittatus (now considered A. quinquevittatus) 12S ribosomal RNA gene. This is 99% identical to Excidobates mysteriosus, and not similar to other Adelphobates quinquevittatus individuals. The inclusion of this sample could explain why Roberts et al. (2006a) found a novel placement of this taxon. 4. AF482794.1: Dendrobates biolat (now considered R. sirensis) 16S ribosomal RNA gene. This sequence is very similar to other sequences of R. fantastica and not very similar to other members of the vanzolinii group or, more importantly, to other R. sirensis. 5. All Brazilian D. ventrimaculatus from Symula et al. (2003) and Roberts et al. (2006a) due to ambiguous labels on both the vial containing received tissue and corresponding DNA extraction, locality cannot be attributed with certainty to either locality/species. At the time of DNA extraction this information may have been clear; however, as outsiders, we cannot determine any locality with certainty.

References Aichinger, M. (1991) A new species of poison-dart frog (Anura: Dendrobatidae) from the Serrania de Sira, Peru. Herpetologica, 47, 1– 5. Almendáriz, L.A. (1987) Contribucion al conocimiento de la herpetofauna centroriental ecuatoriana. Politecnica, 12, 77–133. Altig, R. & McDiarmid, R.W. (1999) Body plan: development and morphology. In: McDiarmid, R.W. & Altig, R. (Eds.) Tadpoles: The Biology of Anuran Larvae. University of Chicago Press, Chicago, pp. 24–51. Applied Biosystems, Inc. (1995) Autoassembler v. 1.4.0. Foster City, CA, USA: AppliedBiosystems, Inc. Ávila-Pires, T.C.S. & Hoogmoed, M.S. (1997) The Herpetofauna. In: Lisboa, P. (ed.), Caxiuanã, Museu Paraense Emílio Goeldi, Belém, pp. 389–401. Barnosky, A.D., Matzke, N., Tomiya, S., Wogan, G.O.U., Swartz, B., Quental, T.B., Marshall, C., McGuire, J.L., Lindsey, E.L., Maguire, K.C., Mersey, B., Ferrer, E.A. (2011). Has the Earth's sixth mass extinction already arrived? Nature, 471 (7336), 51–57. Bauer, L. (1988) Pijlgifkikkers en verwanten: de familie Dendrobatidae. Het Paludarium, 1 Nov, 1988, 1–6. Bechter R. & Lescure, J. (1982) Dendrobates quinquevittatus. Fortpflanzungsverhalten im Terrarium und Vielgestaltigkeit der Art (Teil 1). Herpetofauna, 4 (21), 26–30. Bechter R. & Lescure, J. (1983) Dendrobates quinquevittatus. Fortpflanzungsverhalten im Terrarium und Vielgestaltigkeit der Art (Teil 1). Herpetofauna, 5 (22), 28–32. Bernal-Bautista, M.H., Luna-Mora, V.F., Gallego, O. & Quevedo-Gil, A. (2007) A new species of poison frog (Amphibia: Dendrobatidae) from the Andean mountains of Tolima, Colombia. Zootaxa, 1638, 59–68. Boulenger, G.A. (1884 "1883") On a collection of frogs from Yurimaguas, Huallaga River, Northern Peru. Proceedings of the Zoological Society of London, 1883, 635–638, 1 colour pl, 1 B&W pl. Boulenger, G.A. (1899) Descriptions of new batrachians in the collection of the British Museum (Natural History). Annals and Magazine of Natural History, ser. 7, 3, 273–277, 2 plates. Brandley, M.C., Schmitz, A. & Reeder, T.W. (2005) Partitioned Bayesian Analyses, Partition Choice, and the Phylogenetic Relationships of Scincid Lizards. Systematic Biology, 54, 373–390. Brown, J.L., Chouteau, M., Glenn, T. & Summers, K. (2009a) The development and analysis of twenty-one microsatellite loci for three species of Amazonian poison frogs Conservation Genetics Resources, 1, 149–151. Brown, J.L., Morales, V. & Summers, K. (2008a) Divergence in parental care, habitat selection and larval life history between two species of Peruvian poison frogs: an experimental analysis. Journal of Evolutionary Biology, 21, 1534–1543. Brown, J.L., Morales, V. & Summers, K. (2009b) Home range size and location in relation to reproductive resources in poison frogs (Dendrobatidae): a Monte Carlo approach using GIS data. Animal Behaviour, 77, 547–554. Brown, J.L., Morales, V. & Summers, K. (2009c) Tactical reproductive parasitism via larval cannibalism in Peruvian poison frogs. Biology Letters, 5, 148–151.

BROWN ET AL.

Brown, J.L., Morales, V. & Summers, K. (2010) A Key Ecological Trait Drove the Evolution of Biparental Care and Monogamy in an Amphibian. The American Naturalist, 175, 436–446. Brown, J.L., Schulte, R. & Summers, K. (2006) A new species of Dendrobates (Anura : Dendrobatidae) from the Amazonian lowlands in Peru. Zootaxa, 45–58. Brown, J.L. & Twomey, E. (2009) Complicated histories: three new species of poison frogs of the genus Ameerega (Anura: Dendrobatidae) from north-central Peru. Zootaxa, 2049, 1–38. Brown, J.L., Twomey, E., Morales, V. & Summers, K. (2008b) Phytotelm size in relation to parental care and mating strategies in two species of Peruvian poison frogs. Behaviour, 145, 1139–1165. Brown, J.L., Twomey, E., Pepper, M. & Rodriguez, M.S. (2008c) Revision of the Ranitomeya fantastica species complex with description of two new species from Central Peru (Anura : Dendrobatidae). Zootaxa, 1823, 1–24. Calderhead, B. & Girolami, M. (2009) Estimating Bayes factors via thermodynamic integration and population MCMC. Computer Statistics & Data Analysis, 53, 4028–4045. Caldwell, J. & de Oliveira, V. (1999) Determinants of biparental care in the spotted poison frog, Dendrobates vanzolinii (Anura : Dendrobatidae). Copeia, 565–575. Caldwell, J.P. (1997) Pair bonding in spotted poison frogs. Nature, 385, 211. Caldwell, J.P. & Myers, C.W. (1990) A new poison frog from Amazonian Brazil, with further revision of the quinquevittatus group of Dendrobates American Museum Novitates, 2988, 1–21. Caldwell, J.P., Lima, A.P., & Biavati, G.M. (2002) Descriptions of tadpoles of Colostethus marchesianus and Colostethus caeruleodactylus (Anura: Dendrobatidae) from their type localities. Copeia, 2002, 166–172. Charif, R.A., Waack, A.M. & Strickman, L.M. (2010) Raven Pro 1.4 User’s Manual. Cornell Lab of Ornithology, Ithaca, NY. Christmann, S.P. (2004) Dendrobatidae - Poison Frogs - A Fantastic Journey through Ecuador, Peru and Colombia (Volumes I, II & III). Cisneros-Heredia, D.F. (2003) La herpetofauna de la Estacion de Biodiversidad Tiputini, Provincia de Orellana, Amazonia Ecuador. Mem.1er Congreso Ecuatatoriano de Ecologia and Ambeinte, Univ. San Francisco de Quito, Quito, Ecuador. Clough, M. & Summers, K. (2000) Phylogenetic systematics and biogeography of the poison frogs: evidence from mitochondrial DNA sequences. Biological Journal of the Linnean Society, 70, 515–540. Crump, M.L. (1971) Quantitative analysis of the ecological distribution of a tropical herpetofauna.Occasional Papers of the Museum of Natural History of the University of Kansas, 3, 1–62. Daly, J.W., Myers, C.W. & Whittaker, N. (1987) Further classification on skin alkaloids from neotropical posion frogs (Dendrobatidae), with a general survey of toxic/noxious substances in the amphibia. Toxicon, 25, 1023–1095. Darst, C. & Cannatella, D. (2004) Novel relation-ships among hylold frogs inferred from 12S and 16S mitochondrial DNA sequences. Molecular Phylogenetics and Evolution, 31, 462–475. De La Riva, I., Köhler, J., Lötters, S. & Reichle, S. (2000) Ten years of research on Bolivian amphibians: updated checklist, distribution, taxonomic problems, lityerature and iconography. Revista Española de Herpetología, 14, 19–164. Divossen, H. (1999) Dendrobates fantasticus in the field and in the terrarium. Aquarium (Bornheim), 355, 58–60. Divossen, H. (2000) Dendrobates reticulatus - a small jewel from the rainforest. Draco, 1, 58–61. Divossen, H. (2002) Ein Frosch der so aussieht wie er heißt: Dendrobates fantasticus. Faszination Pfeilgiftfrösche, 1, 20–25. Doan, T.M. & Arriaga, W.A. (2002) Microgeographic variation in species composition of the herpetofaunal communities of Tambopata Region, Peru. Biotropica, 34, 101–117. Dubois., A (1988) The genus in zoology: a contribution to the theory of evolutionary systematics. Mémoires du. Muséum de National d’Histoire Naturelle (A), 140, 1–123. Duellman, W.E. & Thomas, R. (1996) Anuran amphibians from a seasonally dry forest in southeastern Peru and comparisons of the anurans among sites in the Upper Amazon Basin. Occasional Papers of the Natural History Museum, University of Kansas, Lawrence, Kansas, 1–34. Duellman, W.E. (1978) The biology of an equatorial herpetofauna in Amazonian Ecuador. Miscellaneous Publications of the University of Kansas, Museum of Natural History, Lawrence, 1–352. Edgar, R.C. (2004) MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Research, 32(5), 1792–1797. Estupinan-T, R.A., Bernardi, J.A.R. & Galatti, U. (2001) La fauna anura en la floresta Nacional de Caxiuanã. In: Lisboa, P. (ed.), Caxiuanã, Museu Paraense Emílio Goeldi, Belém. Erdtmann, L. & Amézquita, A. (2009) Differential Evolution of Advertisement Call Traits in Dart-Poison Frogs (Anura: Dendrobatidae). Ethology, 115, 801–811. Frost, D.R., Grant, T., Faivovich, J., Bain, R.H., Haas, A., Haddad, C.F.B., De SÁ, R.O., Channing, A., Wilkinson, M., Donnellan, S.C., Raxworthy, C.J., Campbell, J.A., Blotto, B.L., Moler, P., Drewes, R.C., Nussbaum, R.A., Lynch, J.D., Green, D.M. & Wheeler, W.C. (2006) The Aphibian Tree of Life. Bulletin of the American Museum of Natural History, 297, 1–291. Frost, D.R. (2011) Amphibian Species of the World: an Online Reference. Version 5.5 (31 January, 2011). Electronic Database accessible at http://research.amnh.org/vz/herpetology/amphibia/American Museum of Natural, New York, USA. Fujitani, T., Hashikawa, H. & Nishio, K. (1998) Breeding the Amazonian poison frog, Dendrobates ventrimaculatus. Animals and Zoos, 50, 74–77. Goebel, A. M., Donnelly, J. M., & Atz., M.E. (1999) PCR primers and amplification methods for 12S ribosomal DNA, the control region, cytochrome oxidase I, and cytochrome b in bufonids and other frogs, and an overview of PCR primers which have amplified DNA in amphibians successfully. Molecular Phylogenetics and Evolution, 11, 163–199. Gosner, K.L. (1960) A simplified table for staging anuran embryos and larvae with notes on Identification. Herpetologica, 16, 183– 190. Graham, C., Ron, S., Santos, J., Schneider, C. & Moritz, C. (2004) Integrating phylogenetics and environmental niche models to explore speciation mechanisms in dendrobatid frogs. Evolution, 58, 1781–1793. Grant, T., Frost, D.R., Caldwell, J.P., Gagliardo, R., Haddad, C.F.B., Kok, P.J.R., Means, D.B., Noonan, B.P., Schargel, W.E. & Wheeler, W. (2006) Phylogenetic systematics of dart-poison frogs and their relatives (Amphibia, Athesphatanura, Dendrobatidae). Bulletin of the American Museum of Natural History, 299, 1–262.,

TAXONOMIC REVISION OF RANITOMEYA

107

Gorzula, S. (1996) The trade in dendrobatid frogs from 1987 to 1993. Herpetological Review, 27, 116–123. Hauswaldt, J.S., Ludewig, A.-K., Vences, M. & Pröhl (2010) Widespread co-occurrence of divergent mitochondrial haplotype lineages in a Central American species of poison frog (Oophaga pumilio). Journal of Biogeography. Herrmann, H.J. (1988) Der Netzbaumsteiger, Dendrobates reticulatus Boulenger, 1884 – eine Verhaltensskizze. Aquarien Terrarien, 78–79. Huelsenbeck, J.P. & Ronquist, F. (2001) MrBayes: Bayesian inference of phylogeny. Bioinformatics, 17, 754–755. IUCN (2010) IUCN Red List Categories and Criteria: Version 8.1. IUCN Species Survival Commission, IUCN, Gland, Switzerland and Cambridge, UK, 85 pp. Jungfer, K.-H., Lötters, L. & Jörgens, D. (2000) Der kleinste Pfeilgiftfrosch – eine neueDendrobates-Art aus West-Panama. Herpetofauna, 22, 11–18. Kass, R.E. & Raftery, A.E. (1995) Bayes Factors. Journal of the American Statistical Association, 90, 773–795. Kneller, M. (1983) Beobachtungen anDendrobates fantasticus im natürlichen Lebensraum und im Terrarium. Das Aquarium, 16, 148– 151. Kocher, T.D., Thomas, W.K., Meyer, A., Edwards, S.V., Pääbo, S., Villablanca, F.X. & Wilson, A.C. (1989) Dynamics of mitochondrial DNA evolution in animals: amplification and sequencing with conserved primers. Proceedings of the National Academy of Science, USA, 86, 6196–6200. Kok, P.J.R. (2000) A survey of the anuran fauna of Montagne Belvédère, county of Saül, French Guiana: field list with comments on taxonomy and ecology. The British Herpetological Society Bulletin, 71, 6–26 Lartillot, N. & Philippe, H. (2006) Computing Bayes Factors Using Thermodynamic Integration. Syst Biol, 55, 195–207. Lima, A.P., Magnusson, W.E., Menin, M., Erdtmann, L.K., Rodrigues, D.J., Keller, C. & Hödl, W. (2006) Guia de sapos da Reserva Adolpho Ducke - Amazônia Central (Guide to the frogs of Reserva Adolpho Ducke - Central Amazonia). Attema - Manaus. 168 pp Lescure, J. & Bechter, R. (1982) Le comportement de reproduction en captivite et lepolymorphisme de Dendrobates quinquevittatus Steindachner (Amphibia, Anura,Dendrobatidae). Revue Française d'Aquariologie, Herpétologie, 8,107–118. Lescure, J. & Marty, C. (2001) Atlas des Amphibiens de Guyane. Patrimoines Naturels, Paris. Lötters, S. (1987) Redefinition von Dendrobates quinquevittatus Steindachner, 1834. Salamandra, 24, 72–74. Lötters, S., Reichle, S. & Jungfer, K.-H. (2003) Advertisement calls of neotropical poison frogs (Amphibia: Dendrobatidae) of the genera Colostethus, Dendrobates and Epipedobates, with notes on dendrobatid call classification. Journal of Natural History, 37, 1899–1911, Lötters, S., Jungfer, K.-H., Schmidt, W. & Henkel, F.W. (2007) Poison Frogs: Biology, Species and Captive Husbandry Edition Chimaira, Frankfurt am Main, 668 pp. Lötters, S., van der Meijden, A., Rödder, D., Köster, T.E, Kraus, T., La Marca, E., Haddad, C. F. B. & Veith, M (2010) Reinforcing and expanding the predictions of the distubrance vicariance hypothesis in Amazonian harlequin frogs: a molecular phylogenetic and climate envelope modelling approach. Biodiversity and Conservation, 19, 2125–2146. Lötters, S. & Vences, M. (2000) Bemerkungen zur Nomenklatur und Taxonomie peruanischer Pfeilgiftfrösche (Anura: Dendrobatidae: Dendrobates, Epipedobates). Salamandra, 36, 247–260. Maddison, D., & Maddison, W. (2001) MacClade 4: Analysis of Phylogeny and Character Evolution, Version 4.0. Sinauer Associates, Sunderland, Massachusetts, U.S.A. Maldonado, M. M. & Reichle, S. (2007) Primer registro de Ranitomeya biolat (Morales, 1992) (Anura: Dendrobatidae) para Bolivia, con apuntes sobre la nueve nomenclatura de Dendrobatidae. Kempffinan, 3 (1), 14–17. Marty, C. & Gaucher, P. (1999) Guide sonore des Amphibiens de Guyane. Ceba (Centre Bioacoustique Alpin), France. Mayer, E. & Ashlock, P.D. (1991) Principles of Systematic Zoology, 2nd, McGraw-Hill, Inc. New York. McClain, C. (2011) The mass extinction of scientists who study species.IRCF Reptiles & Amphibians, 18 (1), p 39–41. Medina-Müller, M. (2006) Factores que Influyen en el Canibalismo entre Renacuajos de la Rana Venenosa Dendrobates biolat Morales 1998 (Anura: Dendrobatidae). Tesis, Licenciatura. Universidad Ricardo Palma, Lima, Perú. Meede, U. (1980) Beobachtungen an Dendrobates quinquevittatus und Phyllobates fermoralis. Salamandra, 16, 38–51. Melin, D. E. (1941) Contributions to the knowledge of the Amphibia of South America. Göteborgs Kungl. Vetenskaps-och Vitterhetssamhälles.Handlingar. Serien B, Matematiska och Naturvetenskapliga Skrifter, 1, 1–71. Melo-Sampaio, P. R. & Barbosa de Souza, M. (2009) Ranitomeya biolat. Herpetological Review, 40 (4), 447. Morales, V. (1992) Dos especies nuevas de Dendrobates (Anura: Dendrobatidae) para Perú. Caribbean Journal of Science, 28, 191– 199. Moreira, G.R., Gordo, M., Martins, M., Galatti, U. & Oda, W.Y. (1997) Relatório Final da Área Temática Herpetofauna. Macrozoneamento Sócio-Econômico-Ecológico do Estado de Rondônia. Planafloro report, pp. 1–57. Planafloro, Porto Velho. Myers, C., W. (1982) Spotted poison frogs: Description of Three new Dendrobates from Western Amazonia, and resurrection of a lost species from "Chiriqui". American Museum Novitates, 2721, 23. Myers, C.W. & Daly, J.W. (1976) Preliminary evaluation of skin toxins and vocalizations in taxonomic and evolutionary studies of poison-dart frogs (Dendrobatidae). Bulletin of the American Museum of Natural History, 157, 173–262. Myers, C.W. & Daly, J.W. (1980) Taxonomy and ecology of Dendrobates bombetes, a new Andean poison frog with new skin toxins. American Museum Novitates, 1–23. Nijman, V. & Shepherd, C.R. (2010) The role of Asia in the global trade in CITES II-listed poison arrow frogs: hopping from Kazakhstan to Lebanon to Thailand and beyond. Biodiversity & Conservation, 19, 1963–1970. Noonan, B.P. & Wray, K.P. (2006) Neotropical diversification: the effects of a complex history on diversity within the poison frog genus Dendrobates. Journal of Biogeography, 33, 1007–1020. Nylander, J.A. (2004) MrModeltest. Program distributed by the author Nylander, J.A.A., Ronquist, F., Huelsenbeck, J.P. & Nieves-Aldrey, J.L. (2004) Bayesian phylogenetic analysis of combined data. Systematic Biology, 53, 47–67. Palumbi, S., Martin, A., Romano, S., McMillan, W. O., Stice, L., & Grabowski, G.(1991). The simple fools guide to PCR, Version 2.0. University of Hawaii. Pepper, M.P., Twomey, E. & Brown, J.L. (2007) The smuggling crisis. Leaf Litter Magazine.Tree Walkers International, 1.1, 5–7 Perez-Peña, P., Chavez, G., Twomey, E. & Brown, J.L. (2010) Two new species of Ranitomeya (Anura: Dendrobatidae) from eastern

BROWN ET AL.

Amazonian Peru. Zootaxa, 2439, 1–23. Phillips, S.J., Anderson, R.P. & Schapire, R.E. (2006) Maximum entropy modeling of species geographic distributions. Ecol Model, 190, 231–259. Pistoni, J. & Felipe-Toledo L. (2010) Amphibian Illegal Trade In Brazil: What Do We Know? South American Journal of Herpetology, 5(1), 2010, 51–56. Poelman, E.H. & Dicke, M. (2007) Offering offspring as food to cannibals: oviposition strategies of Amazonian poison frogs (Dendrobates ventrimaculatus). Evolutionary Ecology, 21, 215–227. Poelman, E.H. & Dicke, M. (2008) Space use of Amazonian poison frogs: Testing the reproductive resource defense hypothesis. Journal of Herpetology, 42, 270–278. Pröhl, H. (2005) Territorial behavior in Dendrobatid Frogs. Journal of Herpetology, 39, 354–365. Quicke, D. L. (1993) Principles and Techniques of Contemporary Taxonomy. London, Blackie Academic and Professional. Rambaut, A., Drummond, A. J., (2007) Tracer v1.4, Available from:http://beast.bio.ed.ac.uk/Tracer (6/1/2009) Roberts, J.L., Brown, J.L., von May, R., Arizabal, W., Presar, A., Symula, R., Schulte, R. & Summers, K. (2006a) Phylogenetic relationships among poison frogs of the genus Dendrobates (Dendrobatidae): A molecular perspective from increased taxon sampling. Herpetological Journal, 16, 377–385. Roberts, J.L., Brown, J.L., von May, R., Arizabal, W., Schulte, R. & Summers, K. (2006b) Genetic divergence and speciation in lowland and montane peruvian poison frogs. Molecular Phylogenetics and Evolution, 41, 149–164. Rodríguez, L. O. & Duellman, W. E. (1994) Guide to the frogs of the Iquitos Region, Amazonian Peru. University of Kansas, Natural History Museum, Special Publication, 22,1–80. Rodriguez, L.O. & Knell, G. (2003) Amphibians and reptiles. In: Pitman, N., Vriesendorp, C., Moskovits, D. (Eds), Rapid biological inventory: Peru: Yavari. The Field Museum, Chicago, U.S.A., pp. 167–170. Ruiz-Carranza, P. M., & Ramírez-Pinilla, M. P. (1992) Una nueva especies de Minyobates (Anura: Dendrobatidae) de Colombia. Lozania, Bogotá 61, 1–16. Ruxton, G., Sherratt, T. & Speed, M. (2004) Avoiding Attack - The Evolutionary Ecology of Crypsis, Warning Signals and Mimicry. Oxford: Oxford University Press. Santos, J., Coloma, L. & Cannatella, D. (2003) Multiple, recurring origins of aposematism and diet specialization in poison frogs. Proceedings of the National Academy of Sciences of the United States of America, 100, 12792–12797. Santos, J.C., Coloma, L.A., Summers, K., Caldwell, J.P., Ree, R. & Cannatella, D.C. (2009) Amazonian Amphibian Diversity Is Primarily Derived from Late Miocene Andean Lineages. PLoS Biol, 7, e1000056. Schulte, L. M., Rödder, D., Schulte, R. & Lötters S. (2010) Preference and competition for breeding plants in coexisting Ranitomeya species (Dendrobatidae): does height play a role? Salamandra 46(3), 180–184. Schulte, R. (1981) D. quinquevittatus: Ökologie, Haltung und Zucht. Herpetofauna, 3, 10, 24–28. Schulte,Missing R. (1986) Eine neue Dendrobates-Art aus P.A. Ostperu (Amphibia: Salentia: 8, the 11–20. Reference: Silverstone, (1975) A revision of theDendrobatidae). poison-arrowSauria, frogs of genus Dendrobates Schulte,Wagler. R. (1999) Pfeilgiftfrösche "Artenteil Peru". INBICO, Wailblingen, Germany, 294 pp. Natural History Museum of Los Angeles County, 21, 1–55. Shimada, T., Matsui, M., Yambun, P., Lakim, M., & Mohamed, M. (2008) Detection of two cryptic taxa in Meristogenys amoropalmatus (Amphibia, Ranidae) through nuclear and mitochondrial DNA analyses. Zootaxa, 1843,24–34. Shreve, B. (1935) On a new Teiid and Amphibia from Panama, Ecuador, and Paraguay. Occasional Papers of the Boston Society of Natural History, 8, 209–218. Souza, M.B. (2009) Anfíbios: Reserva Extrativista Alto Juruá e Parque Nacional da Serra do Divisor. Centro de Ciências Biológicas e da Natureza Universidade Federal do AcreCampus Universitário. Suchard, M.A., Weiss, R.E. & Sinsheimer, J.S. (2005) Models for Estimating Bayes Factors with Applications to Phylogeny and Tests of Monophyly. Biometrics, 61, 665–673. Summers, K. (1999) The effects of cannibalism on Amazonian poison frog egg and tadpole deposition and survivorship in Heliconia axil pools. Oecolgia, 119, 557–564. Summers, K. & Amos, W. (1997) Behavioral, ecological, and molecular genetic analyses of reproductive strategies in the Amazonian dart-poison frog, Dendrobates ventrimaculatus. Behavioral Ecology, 8, 260–267. Summers, K., Bermingham, E., Weigt, L. & McCafferty, S. (1997a) Phenotypic and genetic divergence in three species of dart-poison frogs with contrasting parental behavior. Journal of Heredity, 88, 8–13. Summers, K., Bermingham, E., Weigt, L., McCafferty, S. & Dahlstrom, L. (1997b) Phenotypic and genetic divergence in three species of dart-poison frogs with contrasting parental behavior. Journal of Heredity, 88, 8–13. Summers, K. & Earn, D.J. (1999) The cost of polygyny and the evolution of female care in poison frogs. Biological Journal of the Linnean Society, 66, 515–538. Summers, K. & McKeon, C.S. (2004) The evolutionary ecology of phytotelmata use in poison frogs. Miscellaneous Publications of the Museum of Zoology of the University of Michigan, 55–73. Summers, K. & Symula, R. (2001) Cannibalism and kin discrimination in tadpoles of the Amazonian poison frog, Dendrobates ventrimaculatus, in the field. Herpetological Journal, 11, 17–21. Summers, K., Weight, L.A., Boag, P. & Bermingham, E. (1999) The evolution of female parental care in poison frogs of the genus Dendrobates: Evidence from mitochondrial DNA sequences. Herpetologica, 55, 254–270. Swofford, D. L. (2004) PAUP*: Phylogenetic Analysis Using Parsimony (* and OtherMethods), Version 4.0b. Sunderland, MA: Sinauer Associates. Symula, R., Schulte, R. & Summers, K. (2001) Molecular phylogenetic evidence for a mimetic radiation in Peruvian poison frogs supports a Mullerian mimicry hypothesis. Proccedings of the Royal Sociey of Lond Series B-Biological Science, 268, 2415–2421. Symula, R., Schulte, R. & Summers, K. (2003) Molecular systematics and phylogeography of Amazonian poison frogs of the genus Dendrobates. Molecular Phylogenetics and Evolution, 26, 452–75.

TAXONOMIC REVISION OF RANITOMEYA

109

Terborgh, J., & Weske, J.S. (1975) The role of competiion in the distribution of Andean birds. Ecology, 56, 562–576. TRAFFIC (2011). TRAFFIC Bulletin Seizures and Prosecutions: Vol. 16 No. 3 (March 1997) to Vol. 23 No. 4 (April 2011) Twomey, E. & Brown, J.L. (2008) Spotted poison frogs: rediscovery of a lost species and a new genus (Anura: Dendrobatidae) from northwestern Peru. Herpetologica, 64, 121–137. Twomey, E. & Brown, J.L. (2009) Another species of Ranitomeya (Anura: Dendrobatidae) from Amazonian Colombia. Zootaxa, 1302, 48–60. Vanzolini, P.E. (1986) Levantamento herpetologico da área do estado de Rondônia sob a influencia da rodovia BR 364. Relatorio de Pesquisa, 1, 1–50. Vences, M., Kosuch, J., Boistel, R., Haddad, C.F.B., La Marca, E., Lötters, S. & Veith, M. (2003b) Convergent evolution of aposematic coloration in neotropical poison frogs: a molecular phylogenetic perspective. Organism Diversity & Evolution, 3, 215–226. Vences, M., Kosuch, J., Lötters, S., Widmer, A., Jungfer, K.H., Kohler, J. & Veith, M. (2000) Phylogeny and classification of poison frogs (Amphibia: dendrobatidae), based on mitochondrial 16S and 12S ribosomal RNA gene sequences. Molecular Phylogenetics and Evolution, 15, 34–40. Venegas, P.J. & von May, R. (2010) Apéndice/Appendix 4: Anfibios y Reptiles/Amphibians and Reptiles. Pp. 282–286, in Gilmore MP, Vriesendorp C, Alverson WS, del Campo A, von May R, Lopez Wong C, Rios Ochoa S (eds) Perú: Maijuna. Rapid Biological and Social Inventories Report 22. The Field Museum, Chicago. Vitt, L.J. & Caldwell, J.P. (1996) Inventário e Ecologia da herpetofauna da Amazônia: Rio Juruá, Porto Walter, Acre, Brasil. Report Proj. NSF Project DEB-9505518 von May, R., Catenazzi, A., Angulo, A., Brown, J.L., Carrillo, J., Chávez, G., Córdova, J.H., Curo, A., Delgado, A., Enciso, M.A., Gutiérrez, R., Lehr, E., Martínez, J.L., Medina-Müller, M., Miranda, A., Neira, D.R., Ochoa, J.A., Quiroz, A.J., Rodríguez, D.A., Rodríguez, L.O., Salas, A.W., Seimon, T., Seimon, A., Siu-Ting, K., Suárez, J., Torres, J. & Twomey, E. (2008a) Current state of conservation knowledge of threatened amphibian species in Peru. Tropical Conservation Science, 1, 376–396. von May, R., Medina–Müller, M., Donnelly, M.A., & Summers, K. (2008b) The tadpole of the bamboo–breeding poison frog Ranitomeya biolat (Anura: Dendrobatidae). Zootaxa, 1857, 66–68. von May, R., Medina-Müller, M., Donnelly, M.A., & Summers, K. (2009a) Breeding-site selection by the poison frog Ranitomeya biolat in Amazonian bamboo forests: an experimental approach. Canadian Journal of Zoology, 87, 453–463. von May, R., Reider, K.E., & Summers, K. (2009b) Effect of body size on intraguild predation between tadpoles of bamboo-breeding poison frogs and predaceous mosquito larvae. Journal of Freshwater Ecology, 24, 431–435. von May, R. & Mueses-Cisneros (2011) Apéndice/Appendix 5: Anfibios y Reptiles/Amphibians and Reptiles. Pp. 306–311, in Pitman, N.C.A., Vriesendorp, C., del Campo, A., et al. (eds) Perú: Yaguas-Cotuhé. Rapid Biological and Social Inventories Report 23. The Field Museum, Chicago. Waldram, M. (2008) Breeding biology of Ranitomeya biolat in the Tambopata region of Amazonian Peru. Journal of Herpetology, 42, 232–237. Werner P., Elle, O., Schulte, L., & Lötters, S. (2011) Home range behaviour in male and female poison frogs in Amazonian Peru (Dendrobatidae: Ranitomeya reticulata). Journal of Natural History, 45 (1–2), 15–27 Werner P. (2010) Habitatnutzung bei Pfeilgiftfröschen im amazonischenTieflandregenwald in Peru. Elaphe, 4, 15–19. Wiley, E. O. (1981) Phylogenetics: The Theory and Practice of Phylogenetic Systematics. New York, Wiley. Wiltenmuth, E.B. & Nishikawa, K.C. (1994) Scaling of feeding kinematics across four species of dendrobatid frogs. American Zoologist, 34, 57A. Winston, J. E. (1999) Describing species: Practical taxonomic procedure for biologists. New York: Columbia University Press. Zimmermann, H. & Zimmermann, E. (1984) Durch Nachzucht erhalten: Baumsteigerfrosche. Dendrobates quinquevittatus und D. reticulatus. Aquarien Magazin, 18, 35–41. Zimmermann, H. & Zimmermann, E. (1985) Care and rearing of Dendrobates quinquevittatus and Dendrobates reticulatus. Nordisk Herpetologisk Forening, 28, 76–90. Zimmermann, H., & Zimmermann, E. (1988) Etho-Taxonomie und zoogeographische Artengruppenbildung bei Pfeilgiftfroschen (Anura: Dendrobatidae). Salamandra, 24, 125–160.

BROWN ET AL.

Appendix I. Species names, collection localities, and GenBank accession numbers (where available) for taxa included in the phylogenetic analysis. Name in Tree

Coordiantes

Genbank Info

current name (former name) #

Locality (Country:Province/ Department/State: locality)

12S

16S

amazonica (ventrimaculata) 1

French Guiana: Maripa

Noonan and Wray 2006

DQ163088

DQ163080

DQ163069

amazonica (ventrimaculata) 2

French Guiana: Maripa

Noonan and Wray 2006

DQ163085

DQ163067

amazonica (ventrimaculata) 3

French Guiana: Maripa

Noonan and Wray 2006

DQ163086

DQ163078

DQ163068

amazonica (ventrimaculata) 4

French Guiana: Pic Matecho

Noonan and Wray 2006

DQ163083

DQ163075

DQ163066

amazonica (ventrimaculata) 5

French Guiana

Roberts et al. 2006

DQ371302

DQ371313

DQ371332

amazonica (ventrimaculata) 6

French Guiana: Saul

Noonan and Wray 2006

DQ163084

DQ163081

DQ163070

amazonica (ventrimaculata) 7

Unknown

Fouquet et al. 2007

EU201083

amazonica (ventrimaculata) 8

Colombia: Amazonas, Leticia, Km 11 (Leticia-Tarapaca)

-4.11228

-69.93964

this study

JN635787

JN635842

JN635915

amazonica (amazonica) 9

Peru: Loreto: Near ACEER camp, north bank of Napo River

Symula et al. 2003

AF482781

AF482796

AF482811

amazonica (ventrimaculata) 10

French Guiana: Guyane

Santos et al. 2009

EU342681

JN635885

amazonica (ventrimaculata) 11

French Guiana: Kaw

Noonan and Wray 2006

DQ163087

DQ163076

DQ163073

amazonica 12

Unknown

Fouquet et al. 2007

EU201084

amazonica 13

French Guiana

Vences et al. 2003

AY263248

amazonica (sp. aff. fantastica) 14

Peru: Loreto: Pevas

this study

JN635788

JN635843

JN635916

amazonica (ventrimaculata) 15

Colombia: Amazonas: Puerto Nariño

this study

JN635878

JN635936

amazonica 16

Peru: Loreto: undisclosed locality 'arena blanca 1'

AUR

this study

JN651251

JN635940

amazonica 17

Peru: Loreto: undisclosed locality 'arena blanca 2'

AUR

this study

JN651255

JN635944

amazonica 18

Peru: Loreto: undisclosed locality 'arena blanca 2'

AUR

this study

JN651252

JN635941

amazonica 19

Peru: Loreto: undisclosed locality 'arena blanca 1'

AUR

this study

JN651254

JN635943

amazonica 20

Peru: Loreto: undisclosed locality 'arena blanca 1'

AUR

this study

JN651249

JN635938

amazonica 21

Peru: Loreto: undisclosed locality 'arena blanca 1'

AUR

this study

JN651271

JN635958

amazonica 22

Peru: Loreto: undisclosed locality 'arena blanca 2'

AUR

this study

JN651272

JN635959

amazonica 23

Peru: Loreto: undisclosed locality 'arena blanca 2'

AUR

this study

JN651273

JN635960

amazonica 24

Peru: Loreto: undisclosed locality 'arena blanca 2'

AUR

this study

JN651274

JN635961

amazonica 25

Peru: Loreto: km 26 on the road from Iquitos to Nauta

-3.96400

-73.40500

this study

JN635785

JN635840

JN635913

amazonica (ventrimaculata) 26

Peru: Loreto: km 41 on the road from Iquitos to Nauta

-4.07500

-73.46200

this study

JN635786

JN635841

JN635914

amazonica 27

Peru: Loreto: km 26 on the road from Iquitos to Nauta

-3.96400

-73.40500

this study

JN635784

JN635839

JN635912

amazonica (amazonica) 28

Peru: Loreto: Almendras

-3.83411

-73.38975

Symula et al. 2003

AF482770

AF482777

AF482792

arborea

Panama

Clough and Summers 2000

AF128611

AF128610

AF128612

auratus

Panama

Clough and Summers 2000

AF128602

AF098745

AF128603

benedicta

Peru: Loreto: nearby Shucushuyacu

-6.03209

-75.85700

Brown et al. 2008

EU736204

EU736219

EU736191

TAXONOMIC REVISION OF RANITOMEYA

Lat

Long

Study

CytB

111

Name in Tree

Coordiantes

Genbank Info

current name (former name) #

Locality (Country:Province/ Department/State: locality)

12S

16S

bicolor

Colombia: Choco

Clough and Summers 2000

AF128578

AF128577

AF128579

bombetes 1

Colombia: Valle del Cauca: Buga, Buga-Buenaventura road, 1610 masl

Santos et al. 2009

EU342668

bombetes 2

Colombia: Quindio: Barbas, Finlandia, Hacienda Lusitania, 1958 masl

4.67000

-75.67000

Santos et al. 2009

EU342669

JN635880

bombetes 3

Colombia

this study

JN635869

captivus (sp. nov.) 1

Ecuador: Zamora Chinchipe: near Panguitza, 870 masl

Santos et al. 2009

EU342666

JN635879

captivus 2

Peru: Amazonas: Rio Santiago near the cofluence of the Rio Marañon

-4.44670

-77.64360

Twomey and Brown 2008

EU325898

EU325900

EU325902

captivus (sp. nov.) 3

See captivus (sp. nov.) 1

Santos et al. 2009

EU342665

captivus 4

See captivus 2

-4.44011

-77.61493

Twomey and Brown 2008

EU325899

EU325901

castaneoticus

Brazil: Para: 101 km South, 15 km East of Santarem

-3.15067

-54.84247

Symula et al. 2003

AF482774

AF482789

AF482804

claudiae 1

Panama: Bocas del Toro, S end of Isla Popa, 1 km E Sumwood Channel

Frost et al. 2009

Q283042

claudiae 2

See claudiae 1

Frost et al. 2009

DQ502024

JN635926

claudiae 3

Panama: Bocal del Toro

Roberts et al. 2006

DQ371304

DQ371315

DQ371334

claudiae 4

Panama: Bocas del Toro: Isla Colon, Bocas del Drago, 11 masl

Santos et al. 2009

EU342671

JN635882

claudiae 5

Panama: Bocas del Toro, Isla Colon, La Gruta

Grant et al. 2006

DQ502027

DQ502457

Colostethus talamancae

Costa Rica

Clough and Summers 2000

AF128587

AF128586

AF128588

cyanovitatta

Peru: Loreto: Rio Blanco Basin near to the Zona Reservada Sierra del Divisor, 206 m elevation

-6.92000

-73.84583

Perez-Peña et al. 2010

HM038419

HM038422

HM038425

toraro (ventrimaculata sensu lato) 1

Brazil: location uncertain (either Acre: Porto Walter or Amazonas: near Manaus)

Roberts et al. 2006

DQ371308

DQ371319

DQ371338

toraro (ventrimaculata sensu lato) 10

Colombia: Amazonas, Leticia, Km 11 (Leticia-Tarapaca)

-4.11228

-69.93964

this study

JN635753

JN635807

JN635889

toraro (ventrimaculata sensu lato) 2

Brazil: location uncertain (either Acre: Porto Walter or Amazonas: near Manaus)

Roberts et al. 2006

DQ371307

DQ371318

DQ371337

toraro (ventrimaculata sensu lato) 3

See toraro 2

Symula et al. 2003

AF482782

AF482797

AF482812

toraro (ventrimaculata sensu lato) 4

Brazil: Amazonas: Castanho, ca. 40 km S Manaus at km 12 on road to Autazes, 40 masl

-3.51451

-59.82809

Santos et al. 2009

EU342676

toraro (ventrimaculata sensu lato) 5

Sames as toraro 4

-3.51451

-59.82809

Grant et al. 2006

DQ502232

DQ502665

toraro (ventrimaculata sensu lato) 6

Brazil: Amazonas: Rio Ituxi at the Scheffer Madeireira; left bank, 110 masl

Grant et al. 2006

DQ502072

DQ502503

toraro (ventrimaculata sensu lato) 7

Brazil: Amazonas: Rio Ituxi, Scheffer Madeireira

-8.76333

-42.99333

Grant et al. 2006

DQ502071

DQ502502

toraro (ventrimaculata sensu lato) 8

Brazil: Amazonas: Rio Ituxi, Scheffer Madeireira

-8.76333

-42.99333

Brown et al. 2006

JN635804

JN635877

JN635934

toraro (ventrimaculata sensu lato) 9

Brazil: Amazonas: "Solimoes"

Noonan and Wray 2006

DQ163089

DQ163079

DQ163074

Lat

Long

Study

CytB

BROWN ET AL.

Name in Tree current name (former name) #

Coordiantes Locality (Country:Province/ Department/State: locality)

Lat

Long

Genbank Info Study

12S

16S

CytB

defleri 1

Colombia: Vaupés: north bank of Mosiro Itájura (an oxbow lake off Río Apaporis, also known as ‘Lago Taraira’), 98 m elevation, approximately 0.2 km SW of Estación Biológica Caparú

-1.07707

-69.51426

this study

JN635834

JN635907

defleri 2

See defleri 1

-1.07707

-69.51426

this study

JN635780

JN635835

JN635908

defleri 3

See defleri 1

-1.07707

-69.51426

this study

JN635779

JN635833

JN635906

defleri 4

Colombia: Vaupés: north bank of Río Caquetá near Puerto Córdoba (a village 16.6 km NW from La Pedrera), 68 m elevation

-1.28100

-69.71954

Twomey and Brown 2009

GU062190

GU062191

GU062192

dorisswansonae 1

Colombia

this study

JN635861

dorisswansonae 2

Colombia

this study

JN635863

fantastica 1

Peru: San Martin: Lower Huallaga Canyon

-6.52791

-75.92121

Brown et al. 2008

EU736200

EU736215

EU736185

fantastica 2

Peru: San Martin: Lower Huallaga Canyon

-6.44761

-75.89488

Brown et al. 2008

AF412469

EU36194

fantastica 3

Peru: Amazonas: Pongo de Manseriche

AUR

this study

JN635793

JN635920

fantastica 4

Peru: Amazonas: Pongo de Manseriche

AUR

this study

JN635794

JN635848

JN635921

fantastica 5

Peru: San Martin: km 26 road from Tarapoto to Yurimaguas

-6.42717

-76.29080

Brown et al. 2008

EU736216

EU736186

fantastica 7

Peru: Loreto: Varadero

-5.68744

-76.41832

Brown et al. 2008

EU736195

EU736208

EU736180

fantastica 7

Peru: San Martin: km 42 road from Tarapoto to Yurimaguas

-6.43058

-76.66098

Symula et al. 2003

AF4112447

AF412475

EU736193

fantastica 8

Peru: San Martin: Bocatoma near Tarapoto

-6.45467

-76.34885

Brown et al. 2008

EU736196

EU736209

EU736181

flavovitatta 1

Peru: Loreto: Tahuayo

-4.35842

-73.18444

this study

JN635858

JN635928

flavovitatta 2

Peru: Loreto: Tahuayo

-4.35842

-73.18444

this study

JN635862

JN635930

flavovitatta 3

Peru: Loreto: Tahuayo

-4.35842

-73.18444

this study

JN635864

flavovitatta 4

Peru: Loreto: Tahuayo

-4.35842

-73.18444

Roberts et. al. 2006

DQ371306

DQ371317

DQ371336

flavovitatta 5

Peru: Loreto: TamshiyacuTahuayo

-4.00908

-73.10076

this study

JN635860

JN635929

fulguritus 1

Unknown

Vences et al. 1999

AF124116

fulguritus 2

Colombia: Choco: Bahia Solano, Sierra Mecana, 260 m

6.25847

-77.35560

Grant et al. 2006

DQ502106

DQ502538

galactonotus 1

Brazil

Roberts et al. 2006

DQ371300

DQ371311

DQ371330

galactonotus 2

Brazil: Amazonas

Santos et al. 2009

EU342641

granulifera

Costa Rica

Clough and Summers 2000

AF128608

AF098749

AF128609

histrionica 1

Ecuador

Clough and Summers 2000

AF128617

AF128616

U70154

histrionica 2

Ecuador

Vences et al. 2000

AF124098

AF124117

AF173766

Hyloxalus nexipus 1

Peru: San Martin: Cainarachi Valley

-6.43000

-76.29000

Santos et al. 2009

EU342714

Hyloxalus nexipus 2

Peru: San Martin: Cataratas Ahuashiyacu, 14 km NE Tarapoto, 730 masl

-6.46554

-76.30888

Santos et al. 2009

EU342713

imitator 1

Peru: San Martin: Central Huallaga Canyon near Chazuta

-6.54000

-76.11000

Symula et al. 2001

AF412448

AF412476

AF412504

imitator 2

Peru: San Martin: Pongo de Cainarachi

-6.28681

-76.23179

Symula et al. 2001

AF412459

AF412487

AF412515

TAXONOMIC REVISION OF RANITOMEYA

113

Name in Tree

Coordiantes

Genbank Info

current name (former name) #

Locality (Country:Province/ Department/State: locality)

leucomelas

Venezuela

Clough and Summers 2000

AF128593

AF124119

AF128594

lugubris 1

Nicaragua: Rio San Juan: Rio San Juan, Near Isla de Diamante (ca. 15 km SE El Castillo on Rio San Juan), 80 masl

10.93400

-84.30000

Grant et al. 2006

DQ502061

DQ502492

lugubris 2

Panama: Bocas del Toro, S end of Isla Popa, 1 km E Sumwood Channel

Grant et al. 2006

DQ283043

DQ502456

minutus 1

Panama: Cocle, El Cope, Parque Nacional General de Division 'Omar Torrijos Herrera

Grant et al. 2006

DQ502168

DQ502603

minutus 2

Panama

Clough and Summers 2000

AF128590

AF128589

MMU70163

mysteriosus 1

Peru: Amazonas: Tissues from captive individuals housed in Santa Rosa from Tupire Reserve

-5.43904

-78.55452

this study

JN635803

JN635866

JN635931

mysteriosus 2

See mysteriosus 1

-5.43904

-78.55452

this study

JN635868

mysteriosus 3

Unknown

Roberts et al. 2006

AF128590

AF128589

MMU70163

pumilio

Panama: Bocas del Toro

Clough and Summers 2000

AF128614

AF128613

U70147

quinquevittatus 1

Brazil: Amazonas: Rio Ituxi at the Madeireira Scheffer; left bank, 110 masl

Santos et al. 2009

EU342644

quinquevittatus 2

Brazil: Rondonia: Rio Formoso, Parque Estadual GuajaraMirim, 90 km N Nova Mamore, 180 masl

Santos et al. 2009

EU342646

quinquevittatus 3

See quinquevittatus 2

Santos et al. 2009

EU342645

quinquevittatus 4

Brazil: Rondonia, Parque Estadual Guajara-Mirim

-10.32144

-64.56331

Grant et al. 2006

DQ502063

DQ502496

quinquevittatus 5

Brazil: Amazonas, Rio Ituxi, Scheffer Madeireira

-8.47939

-65.71656

Grant et al. 2006

DQ502063

DQ502494

quinquevittatus 6

Brazil: Amazonas, Rio Ituxi, Scheffer Madeireira

-8.47939

-65.71656

Grant et al. 2006

DQ502064

DQ502495

quinquevittatus 7

Brazil: Rondonia, Parque Estadual Guajara-Mirim

-10.32144

-64.56331

Grant et al. 2006

DQ502066

DQ502497

quinquevittatus 8

Brazil: Amazonas, Rio Ituxi, Scheffer Madeireira

-8.47939

-65.71656

Grant et al. 2006

DQ502234

DQ502667

quinquevittatus 9

Brazil

Vences et al. 2003

AY263253

reticulata 1

Peru

Darst et al. 2004

AY326029

reticulata 2

Peru

Santos et al. 2003

AY364567

reticulata 3

Peru: Loreto: "Puerto Almendras, 20 km WSW of Iquitos"

-3.82700

-73.37600

Santos et al. 2009

EU342686

JN635900

reticulata 4

Peru: Loreto: km 55 (Puente Itaya) off of Iquitos- Nauta road

-4.22000

-73.48000

Symula et al. 2003

AF482772

AF482787

AF482802

reticulata 5

Unknown

Vences et al. 2003

AY263245

reticulata 6

Peru

Symula et al. 2001

AF412439

AF412467

AF412495

reticulata (D. sp. Itaya) 7

Peru: Loreto: km 66 (across Itaya River going to Nauta) off of Iquitos- Nauta road

-4.27420

-73.51269

Symula et al. 2003

AF482777

AF482792

AF482807

reticulata 8

Peru: Loreto: km 26 (Allpahuayo) off of IquitosNauta road

-3.96700

-73.41700

Grant et al. 2006

DQ502119

reticulata 9

Peru: Loreto: km 13.5 off of Iquitos-Nauta road

-3.89125

-73.34797

Symula et al. 2003

AF482771

AF482786

AF482801

114 路 Zootaxa 3083 漏 2011 Magnolia Press

Lat

Long

Study

12S

16S

CytB

BROWN ET AL.

Name in Tree current name (former name) #

Coordiantes Locality (Country:Province/ Department/State: locality)

Lat

Long

Genbank Info Study

12S

16S

CytB

reticulata 10

Peru: Loreto: UNAP forests off of Rio Napo

this study

JN635756

JN635810

sirensis (lamasi) 1

Peru: Loreto: nearby Contamana

-7.27412

-75.00242

this study

JN635790

JN635845

JN635918

sirensis (lamasi) 2

Peru: Loreto: nearby Contamana

-7.19854

-74.95244

this study

JN635872

sirensis (lamasi) 3

Peru: Loreto: nearby Contamana

this study

JN635774

JN635828

sirensis (lamasi) 4

Peru: San Martin: nearby Ishinga a small village outside of Tocache

-8.33047

-76.56894

this study

JN635849

JN635923

sirensis (lamasi) 5

Peru: Pasco: nearby Palcazu

-9.82086

-74.96331

this study

JN635796

JN635850

sirensis (lamasi) 6

Peru: Pasco: nearby Palcazu

-9.82086

-74.96331

this study

JN635797

JN635851

sirensis (lamasi) 7

Unknown; likely from Peru: Junin: nearby Iscozacin

this study

JN635792

JN635847

JN635919

sirensis (lamasi) 8

Peru: Huánuco: Puetro Inca

-9.36716

-74.93792

this study

JN635798

JN635852

sirensis (lamasi) 9

Peru: Huánuco: Puetro Inca

-9.45182

-74.80820

this study

JN635799

JN635853

sirensis (sirensis) 10

Peru: Huánuco: western versant of Cordillera El Sira, 4 km from type locality

-9.46360

74.81748

this study

JN635771

JN635825

sirensis (sirensis) 11

See sirensis (sirensis) 10

-9.46360

74.81748

this study

JN635772

JN635826

sirensis (lamasi) 12

Unknown

this study

JN635770

JN635824

sirensis (lamasi) 13

Unknown; likely from Peru: Junin: nearby Iscozacin

this study

JN635791

JN635846

sirensis (lamasi) 14

Unknown

this study

JN635752

JN635806

JN635883

sirensis (biolat) 15

Peru: Madre de Dios: Los Amigos Research Center at the confluence of Rios Los Amigos and Madre de Dios

-12.56700

-70.10000

Roberts et al. 2006

AF482779

AF482794

AF482809

sirensis (lamasi) 16

Peru: Huánuco: Tingo Maria, forest near the Universidad Nacional Agraria de la Selva, (exact paratype locality)

-9.30000

-76.00000

Symula et al. 2003

AF482778

AF482793

AF482808

sirensis (lamasi) 17

Peru: "Rio Pachitea"

this study

JN635769

JN635823

sirensis (biolat) 18

See sirensis (biolat) 15

-12.59653

-20.08967

this study

JN635870

JN635932

sirensis (biolat) 19

See sirensis (biolat) 15

-12.59653

-20.08967

this study

JN635874

JN635933

sirensis (lamasi) 20

Peru: Huánuco: nearby Codo del Pozuzo on the road to Pozuzo

-9.73488

-75.51036

this study

JN635802

JN635856

JN635925

sirensis (lamasi) 21

Peru: Loreto: unspecifed locality

this study

JN635776

JN635830

JN635903

sirensis (lamasi) 22

See sirensis (biolat) 16

-9.30000

-76.00000

this study

JN635800

JN635854

JN635924

sirensis (lamasi) 23

Peru: Huánuco: nearby Tingo Maria

-9.31492

-75.99280

this study

JN635801

JN635855

sirensis (lamasi) 24

Peru: Loreto: nearby Contamana

-7.19854

-74.95244

this study

JN635775

JN635829

JN635902

sp. "Quibdo"

Colombia: Choco: Quibdo, La Troje, 50 masl

Santos et al. 2009

EU342670

JN635881

sp. "supata" 1

Colombia: Cundinamarca, Supatá

-5.05000

-74.25000

this study

JN635865

sp. "supata" 2

Colombia: Cundinamarca, Supatá

-5.05000

-74.25000

this study

JN635867

speciosa

Panama

Clough and Summers 2000

AF128596

AF098747

AF128597

steyermarki 1

Venezuela

Roberts et al. 2006

DQ371310

DQ371321

DQ371340

steyermarki 2

Venezuela

Vences et al. 2003

AY263244

summersi 1

Peru: San Martin: Chipaota

-6.57687

-76.08260

Brown et al. 2008

EU736201

EU736217

EU736187

TAXONOMIC REVISION OF RANITOMEYA

115

Name in Tree

Coordiantes

Genbank Info

current name (former name) #

Locality (Country:Province/ Department/State: locality)

summersi 2

Peru: San Martin: Sauce

-6.72453

-76.25318

Brown et al. 2008

EU736202

EU736218

EU736188

summersi 3

Peru: San Martin: Chazuta

-6.54319

-76.11097

this study

JN635789

JN635844

JN635917

sylvatica

Ecuador

Santos et al. 2003

AY364569

AF324041

terribilis 1

Unknown

Grant et al. 2006

DQ502180

DQ502616

terribilis 2

Colombia: Cauca, Quebrada Guangui, 0.5 km above Rio Patia (upper Saija drainage), 100-200 m

Grant et al. 2006

DQ502157

DQ502593

tinctorius

Captive specimen: presumed from French Guiana

Clough and Summers 2000

AF128605

AF128604

AF128606

tolimensis 1

Colombia: Tolima: Municipio de Falan, Cordillera Central of Colombia

-5.01889

-75.04194

this study

JN635857

tolimensis 2

See tolimensis 1

-5.01889

-75.04194

this study

JN635859

uakarii (ventrimaculata sensu lato) 1

Brazil: location uncertain (either Acre: Porto Walter or Amazonas: near Manaus)

Roberts et al. 2006

DQ371301

DQ371312

DQ371331

uakarii (ventrimaculata sensu lato) 2

See uakarii (ventrimaculata sensu lato) 1

Symula et al. 2003

AF482783

AF482798

AF482813

uakarii (ventrimaculata sensu lato) 3

Brazil: Acre: near 5 km N Porto Walker, inland from the Rio Jurua, 200 masl

-8.25867

-72.77697

Grant et al. 2006

DQ502233

DQ502666

uakarii 4

Peru: Madre de Dios: Los Amigos Research Center at the confluence of Rios Los Amigos and Madre de Dios

-12.59653

-20.08967

this study

JN651270

JN635957

uakarii (ventrimaculata sensu lato) 5

See uakarii (ventrimaculata sensu lato) 3

-8.25867

-72.77697

Grant et al. 2006

DQ502070

DQ502501

uakarii (cf. uakarii) 6

Peru: Huánuco: nearby Tournavista

-8.94822

-74.76783

this study

JN635755

JN635809

uakarii (cf. uakarii) 7

Peru: Huánuco: nearby Tournavista

-8.95347

-74.77928

this study

JN635762

JN635816

JN635895

uakarii 8

Peru: Loreto: TamshiyacuTahuayo, Quebrada Blanco

-4.35842

-73.18444

this study

JN635757

JN635811

JN635890

uakarii 9

Peru: Loreto: TamshiyacuTahuayo, near Tamshiyacu village

-4.00908

-73.10075

this study

JN635761

JN635815

JN635894

uakarii 10

Peru: Loreto: Rio Manati

-3.65201

-72.20045

this study

JN635759

JN635813

JN635892

Lat

Long

Study

12S

16S

CytB

uakarii 11

Peru: Loreto: Rio Manati

-3.65201

-72.20045

this study

JN635760

JN635814

JN635893

uakarii 12

Peru: Loreto: TamshiyacuTahuayo, Quebrada Blanco

-4.35842

-73.18444

this study

JN635758

JN635812

JN635891

uakarii 13

Peru: Loreto: TamshiyacuTahuayo, Quebrada Blanco

-4.35842

-73.18444

Roberts et al. 2006

JN635805

DQ371316

DQ371335

vanzolinii 1

Brazil: Acre: Porto Walter

-8.25867

-72.77697

Grant et al. 2006

DQ502067

DQ502498

vanzolinii 2

Brazil: Acre: Porto Walter

-8.25867

-72.77697

Grant et al. 2006

DQ502236

DQ502669

vanzolinii 3

Brazil: Acre: Porto Walter

-8.25867

-72.77697

Grant et al. 2006

DQ502068

DQ502499

vanzolinii 4

Brazil: Acre: near 5 km N Porto Walter, inland from the Rio Jurua, 200 masl

-8.25867

-72.77697

Santos et al. 2009

EU342673

vanzolinii 5

See vanzolinii 4

-8.25867

-72.77697

Santos et al. 2009

EU342674

vanzolinii 6

Brazil: Acre: Porto Walter

-8.25867

-72.77697

Clough and Summers 2000

AF128599

AF128598

AF128600

variabilis (ventrimaculata) 1

Peru: San Martin: Bonilla km 44 on road from Tarapoto to Yurimaguas

-6.21007

-76.27226

Symula et al. 2001

AF412466

AF412522

AF412494

variabilis (variabilis) 2

Peru: San Martin: km 26 road from Tarapoto to Yurimaguas "Cainarachi Valley"

-6.42717

-76.29080

Symula et al. 2001

AF412521

AF412519

AF412491

variabilis (ventrimaculata) 3

See variabilis (ventrimaculata) 1

-6.21007

-76.27226

Symula et al. 2001

AF412465

AF412521

AF412493

BROWN ET AL.

Name in Tree

Coordiantes

Genbank Info

current name (former name) #

Locality (Country:Province/ Department/State: locality)

variabilis (variabilis) 4

See variabilis (variabilis) 2

-6.42717

-76.29080

Symula et al. 2001

AF412520

AF412492

AF412520

variabilis (variabilis) 5

See variabilis (variabilis) 4

-6.42717

-76.29080

Santos et al. 2009

EU342679

variabilis (variabilis) 6

Peru: San Martin: km 27 road from Tarapoto to Yurimaguas "Cainarachi Valley"

-6.43373

-76.28358

Santos et al. 2009

EU342680

JN635884

variabilis (ventrimaculata) 7

Peru: San Martin: km 7 road from Pongo to Barranquita

-6.28681

-76.23179

this study

JN651247

JN635935

variabilis (ventrimaculata) 8

Peru: San Martin: nearby Pelejo

-6.24352

-75.90925

this study

JN651258

JN635947

variabilis (ventrimaculata) 9

See variabilis (variabilis) 3

-6.19754

-76.25838

this study

JN651257

JN635946

variabilis (variabilis) 10

Unknown

Vences et al. 2003

AY263249

variabilis (ventrimaculata) 11

Peru: San Martin: km 7 road from Pongo to Barranquita

-6.28681

-76.23179

this study

JN651260

variabilis (ventrimaculata) 12

Peru: Loreto: nearby Shucushuyacu

-6.03209

-75.85700

this study

JN651248

JN635937

variabilis (ventrimaculata) 13

Peru: Loreto: nearby Shucushuyacu

-6.03209

-75.85700

this study

JN651263

JN635950

variabilis (ventrimaculata) 14

Colombia: Vaupés: north bank of Río Caquetá 2 km NW of La Pedrera

-1.31053

-69.59990

this study

JN635778

JN635832

JN635905

variabilis (ventrimaculata) 15

See variabilis (ventrimaculata) 14

-1.31053

-69.59990

this study

JN635782

JN635837

JN635910

variabilis (ventrimaculata) 16

See variabilis (ventrimaculata) 14

-1.31053

-69.59990

this study

JN635783

JN635838

JN635911

variabilis (ventrimaculata) 17

Colombia: Vaupés: north bank of Río Caquetá 5 km W of La Pedrera

-1.29720

-69.62690

this study

JN635781

JN635836

JN635909

variabilis (ventrimaculata) 18

Peru: Amazonas: nearby Centro de Espiritualidad TUNAANTS Santa María de Nieva

-4.58000

-77.90000

this study

JN635795

JN635922

variabilis (cf. variabilis) 19

Ecuador: Morona Santiago: nearby Macas

this study

JN635927

variabilis (ventrimaculata) 20

Peru: Loreto: nearby Nauta

-4.46300

-73.58200

this study

JN651250

JN635939

variabilis (ventrimaculata) 21

Peru: Loreto: nearby Puente Nauta

this study

JN651264

JN635951

variabilis (ventrimaculata) 22

Peru: Loreto: from across Itaya river going to Nauta

-4.26667

-73.76167

Symula et al. 2003

AF482784

AF482814

variabilis (ventrimaculata) 23

Peru: Loreto: nearby San Lorenzo, south side of Rio Marañon

-4.92631

-76.48445

this study

JN651268

JN635955

variabilis (ventrimaculata) 24

Peru: Amazonas: Cordillera Campanquiz

-4.42767

-77.52502

this study

JN651267

JN635954

variabilis (ventrimaculata) 25

Peru: Loreto: Tahuayo

Santos et al. 2009

EU342678

variabilis (ventrimaculata) 26

Peru: Loreto: Tahuayo

-4.30750

-73.17778

this study

JN635766

JN635820

JN635898

variabilis (cf. variabilis) 27

Ecuador: Morona Santiago: nearby Macas

this study

JN635763

JN635817

JN635962

variabilis (ventrimaculata) 28

Peru: Loreto: Varadero

-5.68742

-76.41825

this study

JN651256

JN635945

variabilis (ventrimaculata) 29

Ecuador: Sucumbios: Pompeya

-0.43000

-76.62000

Symula et al. 2003

AF482780

AF482795

AF482810

variabilis (ventrimaculata) 30

Ecuador: Sucumbios: Pompeya

-0.43000

-76.62000

Summers et al. 1999

AF098746

AF120013

variabilis (ventrimaculata) 31

Ecuador: Sucumbios: Pompeya

-0.43000

-76.62000

Clough and Summers 2000

AF128621

AF128619

AF128620

variabilis (ventrimaculata) 32

Ecuador: Sucumbios, Estacion Cientifica de Universidad Catolica near Reserva Faunistica Cuyabeno, 220 m

0.00000

-76.16700

Grant et al. 2006.

DQ502069

DQ502500

variabilis (ventrimaculata) 33

Colombia: Vaupés: north bank of Río Caquetá 2 km NW of La Pedrera

-1.31053

-69.59990

this study

JN635777

JN635831

JN635904

TAXONOMIC REVISION OF RANITOMEYA

Lat

Long

Study

12S

16S

CytB

117

Name in Tree current name (former name) #

Coordiantes Locality (Country:Province/ Department/State: locality)

12S

16S

variabilis (ventrimaculata) 34

Ecuador: Francisco de Orellana: Parque Nacional YasuniEstacion PUCE, 230 masl

-0.67131

-76.40050

Santos et al. 2003

AY364570

variabilis (ventrimaculata) 35

Ecuador

this study

JN635754

JN635808

variabilis (ventrimaculata) 36

Peru: Loreto: TamshiyacuTahuayo, Quebrada Blanco

-4.35842

-73.18444

this study

JN635767

JN635821

JN635899

variabilis (ventrimaculata) 37

Peru: Loreto: TamshiyacuTahuayo, Quebrada Blanco

-4.35842

-73.18444

this study

JN651259

variabilis (ventrimaculata) 38

Peru: Loreto: nearby San Lorenzo, south side of Rio Marañon

-4.92631

-76.48445

this study

JN651262

JN635949

variabilis (ventrimaculata) 39

Peru: Loreto: Allpahuayo

Symula et al. 2003

AF482776

AF482791

AF482806

variabilis (ventrimaculata) 40

Peru: Loreto: nearby Shamboyacu on the trail to Puesto 15

-6.94224

-76.07806

this study

JN651266

JN635953

variabilis (ventrimaculata) 41

See variabilis (ventrimaculata) 41

-6.94224

-76.07806

this study

JN651269

JN635956

variabilis (ventrimaculata) 42

Peru: Loreto: nearby Contamana

-7.27412

-75.00242

this study

JN651265

JN635952

variabilis (cf. variabilis) 43

Peru: San Martin: nearby Saposoa on the trail to Cataratas Shima

-6.77107

-76.94127

this study

JN635764

JN635818

JN635896

variabilis (cf. variabilis) 44

Peru: San Martin: nearby Shima

-6.89859

-76.83324

this study

JN635765

JN635819

JN635897

variabilis (cf. variabilis) 45

Peru: San Martin: nearby Saposoa on the trail to Cataratas Shima

-6.77107

-76.94127

this study

JN651261

JN635948

variabilis (ventrimaculata) 46

Peru: Amazonas: Cordillera Campanquiz

-4.44670

-77.64360

this study

JN651253

JN635942

ventrimaculata (duellmani) 1

Ecuador: Francisco de Orellana: Parque Nacional YasuniEstacion PUCE, 230 masl

-0.67131

-76.40050

Santos et al. 2009

EU342688

JN635887

ventrimaculata (duellmani) 2

See ventrimaculata (duellmani) 1

-0.67131

-76.40050

Santos et al. 2009

AY364566

JN635888

ventrimaculata (duellmani) 3

Unknown

Vences et al. 2003

AY263246

ventrimaculata (duellmani) 4

See ventrimaculata (duellmani) 1

-0.67131

-76.40050

this study

JN635768

JN635822

ventrimaculata (D. sp. G) 5

Ecuador

Santos et al. 2003

AY364568

ventrimaculata 6

Peru: Loreto: 40 west of Iquitos north of Rio Nanay

-3.84000

-73.62000

this study

JN635773

JN635827

JN635901

ventrimaculata (duellmani) 7

Colombia: Amazonas, Leticia, Km 11 (Leticia-Tarapaca)

-4.11228

-69.93964

Grant et al. 2006

DQ502266

DQ502697

ventrimaculata (D. sp. Kapawi) 8

Ecuador: Pastaza: Kapawi Lodge, 239 masl

-2.54100

-76.85857

Santos et al. 2009

EU342687

JN635886

virolinensis 1

Colombia: Santander: Virolin, Costilla de Fara, 1767 masl

Santos et al. 2009

EU342667

virolinensis 2

See virolinensis 1

this study

JN635873

virolinensis 3

Colombia: Santander: Socorro

6.47000

73.26000

this study

JN635875

virolinensis 4

Colombia: Santander: Virolin, Costilla de Fara, 1767 masl

this study

JN635871

virolinensis 5

Colombia: Santander: Socorro

6.47000

73.26000

this study

JN635876

yavaricola 1

Peru: Loreto:17 km W of Estiron de Ecuador, 120 m elevation

-4.45972

-71.75097

Perez-Peña et al. 2010

HM038420

HM038423

HM038427

yavaricola 2

See yavaricola 1

-4.45972

-71.75097

Perez-Peña et al. 2010

HM038421

HM038424

HM038428

yavaricola 3

See yavaricola 1

-4.45972

-71.75097

Perez-Peña et al. 2010

HM038426

Lat

Long

Genbank Info Study

CytB

BROWN ET AL.

Appendix II. Material Examined and current taxonomic status (museum-catalog number classification at time of examination, current taxonomic classification*). LACM-71972 Dendrobates altobueyensis (Holotype), Andinobates altobueyensis*; LACM-71973 Dendrobates altobueyensis (Paratype), Andinobates altobueyensis*; LACM-71974 Dendrobates altobueyensis (Paratype), Andinobates altobueyensis*; LACM-71975 Dendrobates altobueyensis (Paratype), Andinobates altobueyensis*; LACM-71976 Dendrobates altobueyensis (Paratype), Andinobates altobueyensis*; LACM71977 Dendrobates altobueyensis (Paratype), Andinobates altobueyensis*; LACM-43802 Dendrobates minutus, Andinobates minutus*; LACM43803 Dendrobates minutus, Andinobates minutus*; LACM-43819 Dendrobates minutus, Andinobates minutus*; LACM-43820 Dendrobates minutus, Andinobates minutus*; LACM-43821 Dendrobates minutus, Andinobates minutus*; LACM-43822 Dendrobates minutus, Andinobates minutus*; LACM-43830 Dendrobates minutus, Andinobates minutus*; LACM-43835 Dendrobates minutus, Andinobates minutus*; LACM-43836 Dendrobates minutus, Andinobates minutus*; LACM-43837 Dendrobates minutus, Andinobates minutus*; LACM-43838 Dendrobates minutus, Andinobates minutus*; LACM-43843 Dendrobates minutus, Andinobates minutus*; LACM-43844 Dendrobates minutus, Andinobates minutus*; LACM-43845 Dendrobates minutus, Andinobates minutus*; LACM-43846 Dendrobates minutus, Andinobates minutus*; LACM-43847 Dendrobates minutus, Andinobates minutus*; LACM-43848 Dendrobates minutus, Andinobates minutus*; LACM-43854 Dendrobates minutus, Andinobates minutus*; LACM-43855 Dendrobates minutus, Andinobates minutus*; LACM-43856 Dendrobates minutus, Andinobates minutus*; LACM-43857 Dendrobates minutus, Andinobates minutus*; LACM-43858 Dendrobates minutus, Andinobates minutus*; LACM-43859 Dendrobates minutus, Andinobates minutus*; LACM-43860 Dendrobates minutus, Andinobates minutus*; LACM-43861 Dendrobates minutus, Andinobates minutus*; LACM-43862 Dendrobates minutus, Andinobates minutus*; LACM-43863 Dendrobates minutus, Andinobates minutus*; LACM-43864 Dendrobates minutus, Andinobates minutus*; LACM-43874 Dendrobates minutus, Andinobates minutus*; LACM-43875 Dendrobates minutus, Andinobates minutus*; LACM-43877 Dendrobates minutus, Andinobates minutus*; LACM-43878 Dendrobates minutus, Andinobates minutus*; LACM-45568 Dendrobates minutus, Andinobates minutus*; LACM-45569 Dendrobates minutus, Andinobates minutus*; LACM-61024 Dendrobates minutus, Andinobates minutus*; LACM-61025 Dendrobates minutus, Andinobates minutus*; LACM-61066 Dendrobates minutus, Andinobates minutus*; LACM-61068 Dendrobates minutus, Andinobates minutus*; LACM-71931 Dendrobates minutus, Andinobates minutus*; LACM-71935 Dendrobates minutus, Andinobates minutus*; LACM-71941 Dendrobates minutus, Andinobates minutus*; LACM-71946 Dendrobates minutus, Andinobates minutus*; LACM-71949 Dendrobates minutus, Andinobates minutus*; LACM-71958 Dendrobates minutus, Andinobates minutus*; LACM-71959 Dendrobates minutus, Andinobates minutus*; LACM-71961 Dendrobates minutus, Andinobates minutus*; LACM-44397 Dendrobates minutus ventrimaculatus (Paratype), Ranitomeya ventrimaculata*; LACM-44398 Dendrobates minutus ventrimaculatus (Paratype), Ranitomeya variabilis*; LACM-43880 Dendrobates opisthomelas, Andinobates opisthomelas*; LACM-43882 Dendrobates opisthomelas, Andinobates opisthomelas*; LACM-43891 Dendrobates opisthomelas, Andinobates opisthomelas*; LACM-43894 Dendrobates opisthomelas, Andinobates opisthomelas*; LACM-43898 Dendrobates opisthomelas, Andinobates opisthomelas*; LACM-43899 Dendrobates opisthomelas, Andinobates opisthomelas*; LACM-71964 Dendrobates opisthomelas, Andinobates opisthomelas*; LACM-71965 Dendrobates opisthomelas, Andinobates opisthomelas*; LACM-71968 Dendrobates opisthomelas, Andinobates opisthomelas*; LACM-71969 Dendrobates opisthomelas, Andinobates opisthomelas*; LACM-71970 Dendrobates opisthomelas, Andinobates opisthomelas*; LACM-42305 Dendrobates quinquevittatus, Ranitomeya amazonica*; LACM-42306 Dendrobates quinquevittatus, Ranitomeya amazonica*; LACM-42307 Dendrobates quinquevittatus, Ranitomeya amazonica*; LACM-42308 Dendrobates quinquevittatus, Ranitomeya amazonica*; LACM-72645 Dendrobates quinquevittatus, Ranitomeya variabilis *; LACM-42309 Dendrobates quinquevittatus, Ranitomeya amazonica*; LACM-72062 Dendrobates quinquevittatus, Ranitomeya variabilis*; LACM-72063 Dendrobates quinquevittatus, Ranitomeya variabilis*; LACM-64596 Dendrobates reticulatus, Ranitomeya reticulata*; USNM-306529 Dendrobates biolat (Paratype), Ranitomeya sirensis*; USNM-537561 Dendrobates biolat, Ranitomeya sirensis*; USNM-537562 Dendrobates biolat, Ranitomeya sirensis*; USNM-537563 Dendrobates biolat, Ranitomeya sirensis*; USNM-537564 Dendrobates biolat, Ranitomeya sirensis*; USNM-537565 Dendrobates biolat, Ranitomeya sirensis*; USNM-537557 Dendrobates biolat, Ranitomeya sirensis*; USNM-537558 Dendrobates biolat, Ranitomeya sirensis*; USNM-268842 Dendrobates biolat (Paratype), Ranitomeya sirensis*; USNM-537559 Dendrobates biolat, Ranitomeya sirensis*; USNM-268843 Dendrobates biolat (Paratype), Ranitomeya sirensis*; USNM537560 Dendrobates biolat, Ranitomeya sirensis*; USNM-268844 Dendrobates biolat (Paratype), Ranitomeya sirensis*; USNM-332406 Dendrobates biolat, Ranitomeya sirensis*; USNM-332407 Dendrobates biolat, Ranitomeya sirensis*; USNM-342778 Dendrobates biolat, Ranitomeya sirensis*; USNM-342881 Dendrobates biolat, Ranitomeya sirensis*; USNM-342882 Dendrobates biolat, Ranitomeya sirensis*; USNM268841 Dendrobates biolat (Paratype), Ranitomeya sirensis*; USNM-269059 Dendrobates biolat (Paratype), Ranitomeya sirensis*; USNM-306528 Dendrobates biolat (Paratype), Ranitomeya sirensis*; USNM-306530 Dendrobates biolat (Paratype), Ranitomeya sirensis*; USNM-306531 Dendrobates biolat (Paratype), Ranitomeya sirensis*; USNM-306532 Dendrobates biolat (Paratype), Ranitomeya sirensis*; USNM-306533 Dendrobates biolat (Paratype), Ranitomeya sirensis*; USNM-127197 Dendrobates fantasticus, Ranitomeya benedicta*; USNM-127198 Dendrobates fantasticus, Ranitomeya benedicta*; USNM-127933 Dendrobates fantasticus, Ranitomeya benedicta*; USNM-299726 Dendrobates imitator, Ranitomeya imitator*; USNM-299727 Dendrobates imitator, Ranitomeya imitator*; USNM-299728 Dendrobates imitator, Ranitomeya imitator*; USNM-299729 Dendrobates imitator, Ranitomeya imitator*; USNM-324338 Dendrobates lamasi, Ranitomeya sirensis*; USNM-166905 Dendrobates lamasi, Ranitomeya sirensis*; USNM-166906 Dendrobates lamasi, Ranitomeya sirensis*; USNM-314946 Dendrobates reticulatus, Ranitomeya reticulata*; USNM-331413 Dendrobates reticulatus, Ranitomeya reticulata*; USNM-222373 Dendrobates reticulatus, Ranitomeya reticulata*; USNM-222374 Dendrobates reticulatus, Ranitomeya reticulata*; USNM-166756 Dendrobates vanzolinii (Paratype), Ranitomeya vanzolinii*; USNM-266119 Dendrobates ventrimaculatus, Not specified*; USNM-320762 Dendrobates ventrimaculatus, Not specified*; USNM537566 Dendrobates ventrimaculatus, Not specified*; USNM-127199 Dendrobates ventrimaculatus, Not specified*; USNM-127200 Dendrobates ventrimaculatus, Not specified*; USNM-317194 Dendrobates ventrimaculatus, Not specified*; USNM-317195 Dendrobates ventrimaculatus, Not specified*; USNM-346308 Dendrobates ventrimaculatus, Not specified*; USNM-520916 Dendrobates ventrimaculatus, Not specified*; USNM520918 Dendrobates ventrimaculatus, Not specified*; USNM-520919 Dendrobates ventrimaculatus, Not specified*; USNM-520920 Dendrobates ventrimaculatus, Not specified*; USNM-520917 Dendrobates ventrimaculatus, Not specified*; MCZ-A-19734 Dendrobates minutus ventrimaculatus (Holotype),*; MCZ-A-19684 Dendrobates quinquevittatus (Paratype), Ranitomeya ventrimaculata*; MCZ-A-19685 Dendrobates quinquevittatus (Paratype), Ranitomeya variabilis*; MCZ-A-19686 Dendrobates quinquevittatus (Paratype), Ranitomeya ventrimaculata*; MCZ-A-19687 Dendrobates quinquevittatus (Paratype), Ranitomeya ventrimaculata*; MCZ-A-19688 Dendrobates quinquevittatus (Paratype), Ranitomeya ventrimaculata*; MCZ-A-19689 Dendrobates quinquevittatus (Paratype), Ranitomeya variabilis*; MCZ-A-19690 (currently in LACM as 44397)

TAXONOMIC REVISION OF RANITOMEYA

Zootaxa 3083 漏 2011 Magnolia Press 路

119

Dendrobates quinquevittatus (Paratype), Ranitomeya ventrimaculata*; MCZ-A-19735 Dendrobates quinquevittatus (Paratype), Ranitomeya ventrimaculata*; MCZ-A-19736 Dendrobates quinquevittatus (Paratype), Ranitomeya ventrimaculata*; MCZ-A-19737 Dendrobates quinquevittatus (Paratype), Ranitomeya variabilis*; MCZ-A-19738 Dendrobates quinquevittatus (Paratype), Ranitomeya ventrimaculata*; MCZ-A-19739 Dendrobates quinquevittatus (Paratype), Ranitomeya variabilis*; MCZ-A-19740 Dendrobates quinquevittatus (Paratype), Ranitomeya ventrimaculata*; MCZ-A-19741 (currently in LACM as 44398) Dendrobates quinquevittatus (Paratype), Ranitomeya variabilis*; MCZ-A-24444 Dendrobates quinquevittatus, Ranitomeya sirensis*; MCZ-A-26050 Dendrobates quinquevittatus, Ranitomeya uakarii*; MCZ-A-28061 Dendrobates quinquevittatus, Ranitomeya defleri*; MCZ-A-37625 Dendrobates quinquevittatus, Ranitomeya amazonica*; MCZ-A-89562 Dendrobates quinquevittatus, Cannot ID species- poor quality*; MCZ-A-92502 Dendrobates quinquevittatus, Ranitomeya variabilis*; MCZ-A-92503 Dendrobates quinquevittatus, Ranitomeya variabilis*; MCZ-A-93643 Dendrobates quinquevittatus, Ranitomeya uakarii*; MCZ-A-93644 Dendrobates quinquevittatus, Ranitomeya uakarii*; MCZ-A-93645 Dendrobates quinquevittatus, Ranitomeya uakarii*; MCZ-A-94730 Dendrobates quinquevittatus, Ranitomeya uakarii*; MCZ-A-95697 Dendrobates quinquevittatus, Ranitomeya amazonica*; MCZ-A-96207 Dendrobates quinquevittatus, Ranitomeya amazonica*; MCZ-A-96369 Dendrobates quinquevittatus, Ranitomeya uakarii*; MCZ-A-96760 Dendrobates quinquevittatus, Ranitomeya variabilis*; MCZ-A-97976 Dendrobates quinquevittatus, Ranitomeya variabilis*; MCZ-A-107797 Dendrobates quinquevittatus, Ranitomeya duellmani*; MCZ-A-107798 Dendrobates quinquevittatus, Ranitomeya reticulata*; MCZ-A-107799 Dendrobates quinquevittatus, Ranitomeya ventrimaculata*; MCZ-A-107812 Dendrobates quinquevittatus, Ranitomeya amazonica*; MCZ-A-19684-90, not labeled #1, Ranitomeya ventrimaculata*; MCZ-A-19684-90, not labeled #2, Ranitomeya ventrimaculata*; MPEG-13838 Dendrobates ventrimaculatus (Holotype), Ranitomeya toraro*; MPEG-13839 Dendrobates ventrimaculatus (Paratype), Ranitomeya toraro*; MPEG-13840 Dendrobates ventrimaculatus (Paratype), Ranitomeya toraro*; MPEG-13841 Dendrobates ventrimaculatus (Paratype), Ranitomeya toraro*; MPEG13842 Dendrobates ventrimaculatus (Paratype), Ranitomeya toraro*; OMNH-37438 Dendrobates ventrimaculatus (Paratype), Ranitomeya toraro*; OMNH-37439 Dendrobates ventrimaculatus (Paratype), Ranitomeya toraro*; OMNH-37440 Dendrobates ventrimaculatus (Paratype), Ranitomeya toraro*; OMNH-37441 Dendrobates ventrimaculatus (Paratype), Ranitomeya toraro*; OMNH-37442 Dendrobates ventrimaculatus (Paratype), Ranitomeya toraro*; MPEG-13036 Dendrobates ventrimaculatus (Paratype), Ranitomeya toraro*; MPEG-13037 Dendrobates ventrimaculatus (Paratype), Ranitomeya toraro*; OMNH-36666 Dendrobates ventrimaculatus (Paratype), Ranitomeya toraro*; OMNH-36667 Dendrobates ventrimaculatus (Paratype), Ranitomeya toraro*; OMNH-36062 Dendrobates ventrimaculatus, Ranitomeya uakarii*; MPEG-12394 Dendrobates ventrimaculatus, Ranitomeya uakarii*; MPEG-12395 Dendrobates ventrimaculatus, Ranitomeya uakarii*; MUSM-26957 Ranitomeya benedicta (Holotype), Ranitomeya benedicta*; MUSM-26956 Ranitomeya benedicta (Paratype), Ranitomeya benedicta*; MUSM-26958 Ranitomeya benedicta (Paratype), Ranitomeya benedicta*; MUSM-26959 Ranitomeya benedicta (Paratype), Ranitomeya benedicta*; MUSM-26960 Ranitomeya benedicta (Paratype), Ranitomeya benedicta*; MUSM-26961 Ranitomeya benedicta (Paratype), Ranitomeya benedicta*; MUSM-26962 Ranitomeya benedicta (Paratype), Ranitomeya benedicta*; MUSM-26855 Ranitomeya lamasi, Ranitomeya sirensis*; MUSM-26891 Ranitomeya lamasi, Ranitomeya sirensis*; MUSM-26812 Ranitomeya lamasi, Ranitomeya sirensis*; MUSM-26844 Ranitomeya lamasi, Ranitomeya sirensis*; MUSM-26809 Ranitomeya lamasi, Ranitomeya sirensis*; MUSM-24935 Ranitomeya lamasi, Ranitomeya sirensis*; MUSM-24936 Ranitomeya lamasi, Ranitomeya sirensis*; MUSM-24937 Ranitomeya lamasi, Ranitomeya sirensis*; MUSM-24938 Ranitomeya lamasi, Ranitomeya sirensis*; MUSM-24039 Ranitomeya lamasi, Ranitomeya sirensis*; MUSM-29072 Ranitomeya lamasi, Ranitomeya sirensis*; MUSM-29086 Ranitomeya lamasi, Ranitomeya sirensis*; MUSM-21734 Ranitomeya lamasi, Ranitomeya sirensis*; MUSM-3625 Ranitomeya ventrimaculata, Ranitomeya variabilis*; MUSM-3626 Ranitomeya ventrimaculata, Ranitomeya variabilis*; MUSM-3527 Ranitomeya ventrimaculata, Ranitomeya variabilis*; MUSM-3628 Ranitomeya ventrimaculata, Ranitomeya variabilis*; MUSM-28623 Ranitomeya ventrimaculata, Ranitomeya variabilis*; MUSM-27457 Ranitomeya ventrimaculata, Ranitomeya variabilis*; MUSM-27454 Ranitomeya ventrimaculata, Ranitomeya variabilis*; MUSM-RVM-132 Ranitomeya duellmani, Ranitomeya ventrimaculata*; MUSM-RVM-138 Ranitomeya uakarii, Ranitomeya uakarii*; MUSM-23246 Ranitomeya uakarii (Holotype), Ranitomeya uakarii*; MUSM-23247 Ranitomeya uakarii (Paratype), Ranitomeya uakarii*; MUSM-23248 Ranitomeya uakarii (Paratype), Ranitomeya uakarii*; MUSM-23249 Ranitomeya uakarii (Paratype), Ranitomeya uakarii*; MUSM-23250 Ranitomeya uakarii (Paratype), Ranitomeya uakarii*; MUSM-26994 Ranitomeya summersi (Paratype), Ranitomeya summersi*; MUSM-26991 Ranitomeya summersi (Paratype), Ranitomeya summersi*; MUSM-26992 Ranitomeya summersi (Paratype), Ranitomeya summersi*; MUSM-26993 Ranitomeya summersi (Paratype), Ranitomeya summersi*; MUSM-26949 Ranitomeya summersi (Paratype), Ranitomeya summersi*; MUSM-26967 Ranitomeya summersi (Paratype), Ranitomeya summersi*; NHML-1947.2.15.1 Ranitomeya fantasica (Holotype), Ranitomeya fantasica*; NHML-1947.2.15.2 Ranitomeya fantasica (Paratype), Ranitomeya fantasica*; NHML-1947.2.15.3 Ranitomeya fantasica (Paratype), Ranitomeya fantasica*; NHML-1947.2.15.4 Ranitomeya fantasica (Paratype), Ranitomeya fantasica*; PUJB-JLB08-001 (field number) Ranitomeya ventrimaculata (Paratype), Ranitomeya defleri*; PUJB-JLB08-002 (field number) Ranitomeya ventrimaculata (Paratype), Ranitomeya defleri*; PUJB-JLB08-003 (field number) Ranitomeya ventrimaculata (Paratype), Ranitomeya defleri*; PUJB-JLB08-004 (field number) Ranitomeya ventrimaculata (Paratype), Ranitomeya defleri*; PUJB-JLB08-005 (field number) Ranitomeya ventrimaculata, Ranitomeya variabilis*; PUJB-JLB08-006 (field number) Ranitomeya ventrimaculata, Ranitomeya variabilis*; PUJB-JLB08-007 (field number) Ranitomeya ventrimaculata, Ranitomeya variabilis*; MZUNAP-01-520 Ranitomeya yavaricola (Holotype), Ranitomeya yavaricola*; MZUNAP-01-519 Ranitomeya yavaricola (Paratype), Ranitomeya yavaricola*; MZUNAP-01-518 Ranitomeya yavaricola (Paratype), Ranitomeya yavaricola*; CORBIDI-2266 Ranitomeya cyanovittata (Paratype), Ranitomeya cyanovittata*; CORBIDI-2991 Ranitomeya cyanovittata (Paratype), Ranitomeya cyanovittata*; MUSM-JLB08-011 (field number) Ranitomeya ventrimaculata, Ranitomeya variabilis*; MUSM-JLB08-012 (field number) Ranitomeya ventrimaculata, Ranitomeya variabilis*; MUSM-JLB08013 (field number) Ranitomeya ventrimaculata, Ranitomeya variabilis*; MUSM-JLB08-014 (field number) Ranitomeya ventrimaculata, Ranitomeya variabilis*; MUSM-408 (specimens were on loan to R. Schulte) Dendrobates rubrocephalus (Type), incertae sedis.

120 路 Zootaxa 3083 漏 2011 Magnolia Press

BROWN ET AL.